Sixteen cases of pigeons without a discoverable trace of gonadal tissue have been found and described. Supplementary groups, one of fifteen birds in which one gonad only was abnormally absent, and still another of seventeen cases of nearly suppressed gonads, comprise the principal material obtained and examined.

It is practically certain that this absence of gonads was complete and persistent—not a temporary condition nor one attained shortly before examination of the birds.

Previous observations on the “no gonad” and “one gonad” condition in birds and in other animals are reviewed.

The basis of the gonadless condition is a purely developmental one, in which it is probable that physiological rather than cytological or genetic disturbance is chiefly concerned.

An attempt is made to identify the type of disturbance involved in this developmental anomaly. It is concluded that such failures of gonad development are a consequence of the failure of all “primordial germ cells” to reach or enter the anlage of the germinal ridge.

Some data are available for a test of this view ; they are found to be in conformity with it.

Despite the complete and persistent absence of testicular tissue, some of these birds developed complete and emphatic masculine behaviour ; some additional secondary sex characters observable in these forms were likewise developed to one or another degree.

Since such birds were, according to all the evidence, entirely without interstitial or spermatogenic tissues, it is clear that the thing which is normally accomplished with the aid of the testis incretion may also be accomplished without it. In such cases it is considered that the effective impulse in the determination of the individual’s sex, is capable of effecting the development of at least certain of these “secondary “sex characters. This material offered no opportunity to examine the effects of loss of gonad incretions on plumage characters.

These results, like those of other parts of our studies, lead to the view that the gonad incretions are “controlled” as well as controlling factors in sex development. That they, like the chromosomal or genetic sex factors, rest upon the more basic condition which we elsewhere identify as metabolic level or rate.

Sixteen cases of pigeons without a visible trace of gonadal tissue have been found in the sex studies which have involved the breeding and hybridisation of these birds. Five of these cases were found by Whitman (1919), but these cases were there reported entirely without comment The eleven additional cases observed by the writer are accompanied by observations which seem to establish the special significance of the entire group of cases. Exact data for age and parentage are available for all of the sixteen instances of this rare condition. Some of these birds died as a result of disease; but in ten cases the examination which demonstrated the complete absence of gonads was made either after killing the healthy bird or under conditions or stages in the life history which practically preclude the possibility of the disappearance of any previously existing gonad as the result of disease. Altogether, three groups of facts will be presented which unite to make it wholly probable that no such tissue ever existed at any period in the life cycle of these individuals.

An additional group of fifteen birds in which one gonad was abnormally absent also supplies evidence that the gonadless condition is a developmental rather than a disease phenomenon. This group further demonstrates that the complete suppression of a single gonad occurs in individuals of either sex, in male and female alike. In still another group of seventeen cases it is made clear that various stages exist, short of complete gonad suppression. These two last-named groups contribute some valuable facts since, as will appear later, they place limits upon the types of genetic and physiologic explanations which might otherwise apply to the three groups of cases.

These cases of gonadless birds, and others with traces of gonadal tissue, now offer a basis of fact for the establishment of three points of importance. First, this absence of gonads was complete and persistent—not a temporary condition nor one attained shortly before the examination of the birds. Second, the basis of the gonadless condition is a purely developmental one in which, probably, physiological rather than cytological or genetic change is chiefly concerned. An attempt is made to identify the specific conditions involved in this developmental miscarriage. Third, despite the complete and persistent absence of testicular tissue, some of these birds developed full masculine behaviour and some other of the secondary sex characters observable in these forms. Thoroughgoing male behaviour and masculine desire were strongly-exhibited by “individuals” which were, according to all the evidence, entirely without interstitial and gametogenic tissue.

From the preceding statements it is evident that at some points these cases touch the general sex problem”; and that they have a rather important bearing on much current discussion concerning the rôle of the internal secretions of the gonads. Since more than one subject and a variety of data must be considered here, it has proved impracticable to make full, citations to the literature on all of the topics discussed.

The present paper is one of four publications (or chapters of a volume in preparation) which give a complete summary account of “Data on Gonads” obtained during the thirteen years of our study of sex in doves and pigeons.

In connection with a report dealing with an hermaphrodite Bullfinch, Poll (1909) has described the case of a duck (Netta. rufina) which was entirely without gonads. He stated that from external appearance this bird was ordinarily regarded as a drake, but that close observation proved the male characteristics to be imperfectly developed. When dissected no trace of gonads could be found. Poll then sectioned the entire gonad-renal region of the animal, and careful examination of the sections also showed no trace of gonadal tissue was present.

Somewhat later Geoffrey Smith found and described four cases of pheasant hybrids (probably Phasianus torquatus × P. colchicus) which were without gonads. These birds all had more or less of a left oviduct present; on this account, apparently, Smith regarded them as females.* A part of Smith’s description of these four cases is as follows :—

“These were all young birds, shot probably in their first winter, and dissection revealed an identical condition in all of them, viz., entire absence of any visible ovary or gonad of any kind, and the presence of a very small, narrow oviduct on the left side which had its opening into the cloaca occluded, and the other end narrowing down to a fine filament. I was unable to detect any trace of germinal tissue in any of these birds; otherwise they appeared healthy and normal. All four birds showed male characters in the plumage to slightly varying extents.”

The above cases, together with five previously mentioned cases listed without comment by Whitman, include all of the instances of this condition that we have been able to find in birds. Very few such cases have been found in other higher animals. There are known, however, a very few instances of the absence of one of the two gonads. Since it appears wholly probable that the absence of one gonad arises from a cause similar to that which at other times results in the complete absence of both gonads, and since we have also fourteen cases of this kind to report, some further reference should at once be made to previously reported instances of this abnormality. In the account that follows it is most convenient to make no separation of cases involving the absence of one gonad from those having both gonads absent.

Bond (1914) and others have reported a very few cases of bird hermaphroditism with no right gonad. It should be noted, however, that unless it could be shown that these birds were originally males they do not necessarily represent cases of the abnormal absence of a gonad, because of the quite normal absence of the right ovary in adult female birds. In none of the previously reported cases of avian hermaphroditism was the left gonad absent. The fifteen cases of one absent gonad to be described here, however, all involve the absence of a gonad which should have been present.

Nagel (1897) has described the case of a girl with the left ovary absent. He states that “the complete absence of both ovaries is very rare and coincides with the absence of all genitalia in aborted embyros incapable of life. When one ovary is absent (also very rare) this till now has been found only in connection with a reduction of the uterine horn of the same side; but this reduction is only more or less complete.” In further discussing conditions in the human, Nagel points out (p. 562) that certain pathological conditions (peritonitische Strange) may completely remove an ovary. He notes, however, that in the case of the absent left ovary referred to above, there was present neither acute nor chronic peritonitis.

Hegar (1878) described the case of a woman of 24 years in which the right ovary was entirely absent. Pütz (1889) reported a hermaphrodite hog with an ovo-testis on the left side and no gonad on the right side Jacquet (1895) described a case in Lacerta in which the gonads were entirely wanting, though right and left oviducts were present. The latter were found to open normally into the body cavity anteriorly and into the cloaca posteriorly. Loisel (1901) reported a frog with the right gonad absent and a small pigmented left ovary.

Some data of interest to one point in the problems treated here have been obtained in insects. We refer to the frequent association of gynandromorphism, hermaphroditism, and hybridity with the “one gonad” or the “no gonad” condition. Of particular importance here is Harrison’s (1920) 214 demonstration that all of the female hybrids resulting from the cross, Oporabia dilutata ♂ (30 chromosomes) × O. autumnata ♀ (38 chromosomes), are without gonads. The reciprocal cross apparently yields males only. Goldschmidt (1916) found that from crosses of females of certain European (F) races of moth with males of a Japanese (G) race many of the intersexual females produced have very rudimentary ovaries. Cockayne (1916) also has recorded the case of a gynandromorph hybrid (♂ Amorpha occellatus × ♀ A. populi) which had no gonads. Adlerz (1908) seems to have described two cases of “imperfect lateral gynandromorphous” ants (Anergates atratulus) in which there were no gonads, though male sexual ducts were present on the left side. One of the two is described as “evidently feeling itself to be a male but was treated by the males in the colony as a female.” Morgan and Bridges (1919) described 65 cases of gynandromorph Drosophila, which include at least 10 cases of “rudimentary” gonads, one gonad absent, or of “none found” (two cases). In none of the above mentioned gonadless vertebrates—where the internal secretions of the gonads are alleged to condition sexual desire and behaviour—has the sex behaviour of the individuals been described. For none of these animals apparently—vertebrate or invertebrate—has an attempt been made to identify the cause of failure in gonad development.

Three groups of cases are to be considered. Group 1 : No gonads present. Group 2 : Gonads almost or quite absent. Group 3 : One gonad abnormally absent. The data included in the (sometimes) extended case descriptions given here will later be found necessary to considerations and conclusions which follow in succeeding sections.

Group I : No Gonads (16 Cases)

  • 1. Sub-family hybrid (♂ Zenaidura carolinensis × ♀ Streptopelia risoria). Dead (cause not recorded) at 4.0 months. “No distinct sex glands and such glands probably not present” (Whitman, 1919, table 100.) No females have yet been obtained from this cross.

  • 2. Specific hybrid (♂ Columba guinea × ♀ Common pigeon—C. tabellaria). Dead 19 days after hatching (1.3 months). “Seemed perfectly well day before death. “I could not discover any sexual organs.” (Whitman, 1919, table 63.)

  • 3. Complex sub-family hybrid (♂ Zenaidura-Zenaida, hybr. × ♀ Streptopelia alba). Dead (cause not stated) at 13 months. “I find no testicles, no ovary ; possibly very obscure rudiments of sex-organs exist but I am unable to distinguish anything of that kind.” (Whitman, 1919,-table 106). Another member of this fraternity—in fact the other egg from this same clutch—also developed (Case 4) no gonads.

  • 4. Complex sub-family hybrid (clutch-mate of above, Case 3). Dead (cause not stated) at 14.3 months. “I find nothing that looks like testicles or ovary.” (Whitman, 1919, table 106.)

  • 5. Complex generic hybrid (♂ Turtur orientalis × St. humilis-risoria, hybr. × ♀ St. risoria). Dead at 3.5 months. (Whitman, 1919, table 61.)

  • 6. Complex sub-family hybrid (♂ complex generic hybr.—from Zenaidura-Zenaida crosses—× ♀ St. risoria-alba, hybr.). Dead extreme pulmonary tuberculosis, at 50.3 months. (Whitman, 1919, table 107; autopsy by O. R.)

  • 7. Generic hybrid (♂ St. risoria × ♀ T. orientalis). Probably killed fighting males, at 215 months old, but reduced in weight and some tuberculosis in left lung. Always behaved as a male (though five years before death it failed to mate with a female hybrid). A small thin left oviduct (wt, 0.0382 gm.) was the only reproductive organ present (Further statement in text)

  • 8. Generic hybrid (♂ T. orientalis × ♀ St. alba). Dead, probably starved as a result of crippled legs, at 2.0 months. Case 12 is from the same father.

  • 9. Generic hybrid (♂ St. Alba × ♀ T. orientalis). Dead 2 days after hatching (0.5 month). Same fraternity as Case 14; Case 47—”one gonad”—is a half-brother. Two half-brothers had left oviducts and another had a left testis one-fiftieth size of right.

  • 10. Specific hybrid (♂ St. alba risoria, hybr., mated to sister). Died hatching, 15-day embryo. No gonads, but in region of gonads a “fatty tissue” noted; this region probably diseased.

  • 11. Sub-family hybrid (♂ Zenaida-Zenaidura, hybr. × ♀ St. risoria). Killed healthy at 10.9 months. Same fraternity as Cases 18 and 19.

  • 12. Generic hybrid (♂ T. orientalisSt. alba). Killed, at 183 months; healthy except for enlarged liver and a well walled-off tubercle on right wing joint No gonads, ducts, or Wolffian bodies, but region clear and healthy. When 4 years old it mated as a male with a sister; nested twice while unmated at 4.5 years old. At 5 to 8 years mated as a male with another sister (see further description in text). Cases 8 and 12 were offspring of the same father (different mothers).

  • 13. Specific hybrid (♂ St. alba-risoria × sister). A 14-day dead embryo (Record by Dr E. H. Behre). Same fraternity as “one gonad,” Case 45.

  • 14. Generic hybrid (♂ St. alba × ♀ T. orientalis). Dead 2 days after hatching (0.5 month). Same fraternity as Case 9; Case 47 is a half-brother.

  • 15. Spilopelia suratensis. A dwarf (undersized, 117 to 128 gms.) bird dead at 26.7 months; Ascaridia present in intestine and spleen enlarged. Failed to mate with a ♀ suratensis (see further statement in text). A fairly developed left oviduct. From second egg of a late clutch after and during “reproductive overwork.” A sister was “one gonad,” Case 46; “no gonad,” Case 16, was a relative (niece or nephew).

  • 16. Spilopelia suratensis. The mother was “one gonad,” Case 46—right waxy only. This bird also related to “no gonad,” Case 15 above. Dead 11 days after hatching (0.9 month). It hatched from the’ fourth egg in life of mother and from the smallest (6.21 gms.) of 85 eggs produced by this parent It was from the first egg produced after a male mate was given the mother (Case 46). Only 33 of these 82 (tested) eggs were fertile in the mating of this mother with one of her own species. (a male imported from Borneo), though 24 of the first 26 produced were fertile (and 17 of the first 20 tested eggs were hatched). All young of this fraternity were short-lived ; 6 died as embryos, 9 died within 18 days of hatching, 16 died at 1.8 to 6.7 months, and for 7 of these the cause of death could not be learned at autopsy (one of them moulted at 14 to 17 days after hatching, was unable to renew its feathers and died at 3.1 months); two other young lived to 9.2 and 14.5 months. The bird which lived 14.5 months was a female which failed to mate with “no gonad,” Case 15, and probably also failed with a female white ring dove. The fraternity to which this case (No. 16) belongs was composed of 11 males, 14 females; one had no gonads, and another i4-day embryo had two plain gonads of abnormal and indecisive nature. From a cross with a white ring dove two additional males and one female were obtained. The mother of this fraternity (“one gonad,” Case No. 46) died of tuberculosis 28 months after Case 16 was hatched, and the record (by Dr E. H. Behre) notes that her “ovary is entirely on the right side— no left ovary.” (Whether the right oviduct was the functional duct in this case was unfortunately not recorded.)

Group 2: Gonads Almost or Quite Absent (17 Cases)

  • 17.

    Sub-family hybrid (♂ Z. carolinensis × St. risoria). Dead (cause not recorded) at 7.0 months. No trace of gonads found. This was the first case of “no gonads” observed by Whitman (1919, table 101), and he was then (1900) unwilling positively to declare them completely absent (No females have yet been obtained from this cross.)

  • 18.

    Complex sub-family hybrid (♂ Zenaida - Zenaidura hybr. × ♀ St. risoria). A sister to Cases n, 19. Dead, tuberculous, at 28.5 months. The bird was at first considered a female; there are, however, no definite behaviour records. (See further statement in text) Though tested with a male of her kind, she laid no eggs. Females do occur when Zenaida-Zenaidura hybrids are crossed with Streptopelia).

    • Right: none; no sexual duct

    • Left: mere trace of immature (embryonic) ovary; very small oviduct

  • 19.

    Complex sub-family hybrid (same fraternity as Cases 11 and 18). Killed (some tuberculosis found in spleen and liver) at 22.6 months. This bird readily won a female mate; it always showed the sex behaviour of a male and nested eggs perfectly during the daylight hours as does a male. In 17 tests of eggs produced by a related female the infertility was complete (this, however, not wholly conclusive that it produced no sperm).

    • Right: none.

    • Left: mere trace of gonad of unknown nature.

  • 20.

    Complex generic hybrid (T. orientalis × ris.-orient. hybr.). Dead (cause not stated) at 41 months. “This bird has no plain sexual organs—just a trace (perhaps) of one testicle. It had behaved as a very pugnacious male trying hard to win a mate. Mated with a ring but no result except he played his part perfectly in sitting, etc. … It seemed a well-formed bird except that the head was noticeably enlarged at the occiput.” (Whitman, 1919, table 51.)

  • 21.

    Complex generic hybrid (♂ St. risoria × ♀ alba-orientalis hybr.). Abnormal dead at 0.9 month. An excessively large occiput and a twisted neck. “Possibly a trace of ovary-like tissue (left) present.” (Whitman, table 50.)

  • 22.

    Complex generic hybrid (♂ St. risoria × ♀ alba-orientalis hybr.). Abnormal young (occiput abnormally large), dead 1 day after hatching (0.5 month). “A questionable trace of (left) ovary here.” (Whitman, 1919, table 50.)

  • 23.

    Complex generic hybrid (alba-orientalis hybr. × ♀alba-risoria hybr.). Parents same as for Case 42. This egg partly opened at end of 15 days’ incubation (1 day after time to hatch) and noted a live embryo. Next day a dead, small, immature (12-day stage) abnormal embryo was found. No upper beak; eyes only 3 mm. diameter (a control = 8.7 mm.). Could find no gonads, but conditions of observation do not exclude the existence of abnormally small traces of such tissue.

  • 24.

    Complex generic hybrid (♂ orient.-turtur-ris.-alba hybr. × ♀ St. risoria). Dead at 75.1 months. Previously thought to be a male though it had wide pubic bones. No distinct ovary but a very few scattered ova at site of left ovary.

  • 25.

    Complex generic hybrid (♂ orient.-ris.-alba hybr. × sister). Killed at 39.8 months with a large tuberculous (?) tumour in region of right gonad; this tumour possibly consumed a former gonad. No gonad present on either side, though both adrenals were entirely plain and distinct.

  • 26.

    Complex trispecific hybrid (♂ St. alba-douraca-risoria hybr. × ♀ St. risoria). Hatched from an “alcoholised” egg (1914 series). Killed, very tubercular at 34.3 months. Began laying when 5.7 months old; laid 48 eggs (17 while mated to sister) 11 of which were thin-shelled, and of her 28 clutches 8 were of “single” eggs. While mated with a healthy male all of the 9 eggs tested were fertile; 2 ♂ and 2 ♀ were hatched. She had entirely ceased laying 9 months before she was killed for examination. Mesentery and spleen involved in tubercles, and if any trace of a gonad existed it was involved in the cheesy mass reaching to the left gonad region. The preceding history of course proves the complete or nearly complete removal of an ovary under this disease

  • 27.

    Trispecific hybrid (♂ St. douraca × ♀ St. alba-risoria). Two members of same fraternity had two unlike gonads of questionable nature. Case 27 died suddenly (breathing difficulties, thought healthy till day of death) at 8.8 months. A large bird at death (wt = 175 gms.). It had been thought a “probable male” (pubic bones only 7.5 mm. apart). (The clutch-mate of this bird died at hatching and had two gonads—left larger—of wholly indecisive appearance)

    • Right: none.

    • Left: ovary consisting of one ovum (4.0 mm. diam.) and a trace (7 mgms.) of other immature ovarian tissue.

  • 28.

    Complex trispecific hybrid (♂ St, alba-douraca-risoria hybr. × ♀ similar). At 2.2 days began subjecting this embryo to low temperatures (11.0° to 18.0°C.) during 11 to 13 hours daily (for 10 days). The 13-day embryo was dead when examined. The gonads were nearly absent and wholly abnormal.

    • Right: an atypical trace

    • Left: a long thin narrow strand or thread (broader anteriorly) of quite uncertain nature.

  • 29.

    Specific hybrid (♂ St. alba-risoria × sister). A 12-day dead white abnormal embryo. Lower beak 2.0 mm. too short. White colour here confined to female sex (16 cases); so, apparently this bird should have developed ovaries.

    • Right: absent, or abnormally minute.

    • Left: absent, or abnormally minute.

  • 30.

    Specific hybrid (♂ St alba-risoria × ♀ St. risoria). A white young dead 1 day after hatching (0.5 month.). All white young of this fraternity were females (46 cases); apparently, therefore, this bird should have had a left ovary.

    • Right: probably none.

    • Left: a minute gonad (barely visible).

  • 31.

    Complex family hybrid (♂ orientalis-risoria hybr. × ♀ common pigeon). Dead at 6.4 months. “I could find neither male nor female organs.” (Whitman, 1919, table 118.)

  • 32.

    Complex family hybrid (♂ Columba-Streptopelia hybr. × ♀ Streptopelia hybr.). Killed, very tuberculous, at 20.8 months. (This bird had only one brother and one sister. The brother was killed healthy at 22 months and had testes with quite abnormal size relations—the right, 0.300 gm., the left, 1.320 gms. (Case C of another list) The sister was killed healthy at 22.5 months and had two very immature ovaries—right smaller; these as little developed as the ovary of a dove two weeks after hatching.)

    • Right: a gonad resembling a persisting right ovary in a bird one week after hatching. No oviduct

    • Left: a little membranous tissue, undeveloped and resembling ovary of a dove 10 days after hatching. Oviduct was large and evidently diseased.

  • 33.

    Generic hybrid (♂ St. alba-risoria × ♀ T. orientalis). Dead at 65.4 months; tuberculosis found in intestines only. This female laid only 17 eggs during her life. She produced her first egg at a very advanced age (19.3 months), though she mated in her first year. Her first egg laid while mated to a female ring dove; otherwise wholly refused to lay eggs when mated to females. Mated to a male during 7 months she laid 16 of her 17 eggs and otherwise showed typical female behaviour, until (onset of tuberculosis) near end of life. At 2.4 months before death (from tuberculosis), however, she proved much more masculine in sex behaviour than her sister mate (during this period she took the part of a male in 115 copulations; the female part in 8 copulations.)

    • Right: none

    • Left: ovary minute, about 0.007 gm.

Group 3: One Gonad Abnormally Absent (15 Cases)

  • 34.

    Turtur orientalis. Dead at 39.5 months; extremely tubercular lungs, soft liver (with haemorrhages); spleen and a left gonad—if this gonad was ever present—involved in a tubercle weighing 3.68 gms. Externally this tubercle seemed double, but was firmly connected and continuous; all tissue except thin covering capsule apparently destroyed. It is not probable, therefore, that this is like other following cases of “one gonad” absent. This bird had fertilised 6 eggs (two with one ♀ St. alba gave 2-day and 4-day embryos; these fertilised when Case 34 was only 312months old. With another similar female 4 young were hatched.) (See text.)

    • Right: apparently pulled over to middle of body; minute, 0.004 gm.

    • Left: if ever present, involved in tumour (tuberculous) with spleen.

  • 35.

    Common pigeon (parents both “scragglies”). Full term embryo (18 days) died hatching.

    • Right: almost certainly no gonad.

    • Left: very small ovary.

  • 36.

    Common pigeon (parents normal, but carried scraggliness). A dead i4-day monster; body very short, legs turned inward. Three of four young of this fraternity which lived long enough to develop feathers had “scraggly” plumage.

    • Right: none.

    • Left: of uncertain nature.

  • 37.

    Common pigeon (parents carried both ataxia and scraggliness). A badly nourished young dead 14 days after hatching (1.1 month old).

    • Right: none.

    • Left: tabulated, abnormal, but apparently a testis.

  • 38.

    Common pigeon. Killed healthy at 11.3 months.

    • Right: none; apparently no duct.

    • Left: abnormal gonad with fluid-filled vesicle; weight (less fluid), 0.020 gm. Duct absent or minute.

  • 39.

    Complex generic hybrid (♂ alba-orientalis, hybr. × ♀ St. alba-douraca-ris., hybr.). An embryo dead at 10 to 11 days after 72 hours’ exposure to 26.8 per cent, oxygen when 1 to 4 days old. Abnormal beak. (A younger clutch-mate sister similarly treated, from 0.2 to 3.2 days, had normal gonads. One male (of other parentage) of same age (1.2 day) survived the same treatment and had normal gonads. But of 9 embryos—brother or sisters—aged 0.2 to 6.2 days at beginning of treatment, 4 (including the present case) had one or both gonads atypical).

    • Right: not present.

    • Left: plainly more like ovary.

  • 40.

    Complex generic hybrid (♂ Stigmatopelia senegalensis - St. alba hybr. × ♀ St. risoria). Dead at 19 months.

    • Right: an abnormal gonad (ovary ?) apparently the seat of a tumour.

    • Left: none.

  • 41.

    Complex generic hybrid (♂ ris-alba-douraea-orient., hybr. × ♀ of same type). Killed healthy at 9.7 months. Had wide pubic bones and until killed was thought to be a female.

    • Right: three or four thin strands radiating backward from anterior normal point of attachment of testicle; these seem (on sectioning) to be formed of cortical suprarenal tissue, not testis (only case ever seen).

    • Left: a very elongate testis (20.3 mm., wt, 0.337 gm.).

  • 42.

    Complex generic hybrid (parents same as for Case 23). A dead full-term embryo (14 days).

    • Right: no gonad.

    • Left: atypical, perhaps more like testis.

  • 43.

    Complex generic hybrid (♂ Zenaida-Zenaidura complex hybr. × ♀ of same). A 13-day dead embryo.

    • Right: a testis-like gonad.

    • Left: no gonad (record by Dr E. H. Behre).

  • 44.

    Complex generic hybrid (♂ alba-orientalis, hybr. × ♀ alba-douraca-ris., hybr.). Killed, healthy, at 1.1 month.

    • Right: no gonad or a trace of gonad in form of a globule opposite anterior one-tenth of left ; possibly a right sperm duct

    • Left: normal testis; questionable duct (record by Dr E. H. Behre).

  • 45.

    Specific hybrid (parents same as for Case 13). A live 12.5-day embryo.

    • Right : none (record by Dr E. H. Behre).

    • Left: ovary.

  • 46.

    Spilopelia suratensis. (Sister to Case 15, “no gonad”; mother of Case 16, “no gonad.”) Dead, tuberculous, at 37.3 months.

    • Right: ovary (had functioned).

    • Left: no gonad (record by Dr E. H. Behre).

  • 47.

    Turtur orientalis (mother was same as for “no gonad,” Cases 9 and 14). Dead at 0.8 month.

    • Right: more like testis.

    • Left: probably not a trace present

  • 48.

    Ocyphaps lophotes. Dead (poor care) 12 days after hatching (0.9 month).

    • Right: a minute globule of uncertain nature; quite abnormal.

    • Left: larger, typical testis.

Distribution of the Occurrence of Various Grades of Gonad Absence

It is essential to any consideration of the present data, and to an interpretation of the matter contained in three additional papers which summarise our data on gonads, that we present the evidence that the various grades of gonad absence (Groups 1 to 3) result from one essentially similar cause. Some data not already considered bear upon this point. This additional material as well as the above-named topic will next be considered.

In connection with the previously given description of the forty-eight individual cases, the occurrence of both similar and different grades of abnormality in other individuals of the same fraternity has been pointed out. The single fact that these cases are drawn from perhaps 20,000 individuals, and probably more than 1000 fraternities, converts this frequency of overlap (of the three grades within single fraternities) into the most cogent evidence that the three grades of gonad absence proceed from the same or a very similar cause. Even within Group 1 (no gonads) six of the sixteen cases belong to two fraternities, and two others are close relatives. One of these sixteen cases has two of its fraternity represented in Group 2 (almost no gonads) ; and three had members of their fraternity in Group 3 (one gonad only).

Here, too, attention may be called to the significant fact that a high proportion of the individuals of all these three groups showed other obvious developmental abnormalities—usually in the form of “bulging occiputs,” “monsters,” “cross-beak,” etc. Group 1 includes two such obvious cases, and the early or sudden death in six others may or may not represent less obvious cases of the same thing. Group 2 includes five notable cases, and Group 3 at least two others, of associated somatic malformation or developmental defect.

Reference to Tables I. and II. will sufficiently demonstrate that the gonad absence or reduction represented in Groups 1 and 2 (at least) are associated with hybridity. In pure species, and in the crosses of the most closely related species, fewest cases result; in the “wide crosses” disproportionately high frequency of gonad absence or reduction is found. Table I. further shows that twelve of the sixteen cases of “no gonads” were the result of a first cross (i.e. were found in F1 hybrids). The cases of Groups 2 and 3, however, were actually most frequent in “second crosses” (usually in offspring of F1 back-crossed or out-crossed, since F1 generics are usually infertile inter se), although the total progeny from these crosses is the smallest (3 per cent.) of the classes represented on the table. Concerning the cases appearing in pure species it should be said that these young were in most, perhaps in all, cases from eggs produced under unusual reproductive conditions—these wild species being subjected to heavy reproductive overwork. Further, it is notable that the relatively light hybridisation involved (related species) in common pigeons and in Streptopelia crosses leave these groups with unfilled quotas of gonad absence. In fact the common pigeon, other than those bearing the recessive ataxia and scraggliness (both plain developmental defects), do not show a single unequivocal case.

Table I.

Three Grades of Gonad Suppression and the Relative Frequency of their Occurrence in Various Stages of Crossing.

The percentages given—in all headings—represent approximately the fraction of the total pigeon population involved in each type of cross.

Three Grades of Gonad Suppression and the Relative Frequency of their Occurrence in Various Stages of Crossing.
Three Grades of Gonad Suppression and the Relative Frequency of their Occurrence in Various Stages of Crossing.
Table II.

Gonad Absence in Relation to Crossing and to “Width of Cross.”

Gonad Absence in Relation to Crossing and to “Width of Cross.”
Gonad Absence in Relation to Crossing and to “Width of Cross.”

Table II. shows that the greatest number of cases in all of the three grades of gonad absence are found in the offspring of generic crosses. The highest ratio of defectives to normals in any of the classes would fall, however, to the sub-family hybrids. But this does not mean that family hybrids would not show a still higher proportion if it were there as often possible to obtain fertile “second crosses”—in which grade of cross the types of defect represented in Groups 2 and 3 are most readily obtained. In another paper dealing with “hermaphrodite and questionable gonads” found in our studies, it will be shown that the greatest number and the highest proportion of such gonads have also occurred in generic hybrids (family and sub-family, o; generic, 14; specific, 3; Streptopelia crosses, 10; all common pigeons, 8; pure species, 6). These figures exclude all cases of all types of so-called “pseudo-hermaphrodites” (i.e. those not involving the gonads). The distribution of these cases among “first crosses” (2 cases), “second crosses” (7 + 2? cases), “later crosses” (6 + 2 ? cases) is also notably similar to the distribution of the cases of Groups 2 and 3. In these two respects, therefore, the occurrence or distribution of the “hermaphroditic and questionable gonads” parallels the two groups of greatly diminished but not completely absent gonads.

Attention should here be drawn to the fact that both males and females show the abnormal absence of one gonad. Obviously it is only in Group 3 that an easy sexual classification of any of our three groups of individuals is possible Within Group 3, however, we are able to show further that the right gonad is sometimes absent in both males (Cases 41, 44, 40) and females (Cases 35, 39, 45) ; and that the left gonad may be absent in either male (Cases 34—and probably 43 and 47) or female (Case 46). Of course we do not mean to assert that the missing gonad, if present, would necessarily have been of the same type as the one found ; nor that even both were present and of the same type, that they would necessarily rightly declare the zygotic composition of the individual Here as elsewhere we do imply that the nature of the gonad—testis or ovary—does satisfactorily declare the actual sex to which the individual belongs, though its zygotic composition may or may not correspond to other members of its sex.

Minor Form Changes in Gonads induced by Changes of Metabolic Rate

The further presentation of evidence that the various degrees of gonad absence are basically of developmental origin (as opposed to disease), and that their solution lies rather in the field of physiology and embryology than in genetics proper, makes it advantageous to utilise here some data on form changes that were experimentally induced in the gonads of early embryos. If embryos be subjected to conditions which temporarily (usually twenty-four to forty-eight hours) change their metabolic rate and, as a result, any notable changes in the form and normal appearance (size, changes, and changes in relative size of right and left gonads are here disregarded) of the gonads are found, this fact would be of interest here. Such effects seem to have been obtained. I have earlier reported (1920, 1923) upon one or another aspect of the result of temporarily subjecting 2100 embryos to the effects of altered oxygen pressures and to low temperatures. Without here considering any details of these experiments we may utilise a fragment of the data thus obtained.

Table III. shows that the embryos which survived the various experimental procedures, and then lived to ages permitting the examination of their gonads, included unusually high percentages of individuals with atypical gonads. A few cases would have been found even without treatment. Certain points in the distribution of these gonad irregularities may be noted incidentally. The atypical condition tabulated more often arises, or is at least more readily observed, in testes than in ovaries. The reduction of metabolism (decreased oxygen ; low temperatures) in the embryo possibly leads to somewhat fewer atypical ovaries, and perhaps to rather more plentiful atypical testes, than does an increased metabolism (increased oxygen). It is also under those conditions which reduce the metabolism that more instances of quite abnormal gonads (rendering the sex unrecognisable) were found.* These details are of limited value because of the small numbers in each class, but they appear to be in satisfactory agreement with our other results on the relation of metabolism to sex.

Table III.

Summary of Numbers and Percentages of distinctly Atypical (not including size, or size relations’) Gonads from Embryos subjected to Altered Oxygen Pressures and Low Temperatures.

Summary of Numbers and Percentages of distinctly Atypical (not including size, or size relations’) Gonads from Embryos subjected to Altered Oxygen Pressures and Low Temperatures.
Summary of Numbers and Percentages of distinctly Atypical (not including size, or size relations’) Gonads from Embryos subjected to Altered Oxygen Pressures and Low Temperatures.

The fact that the above described modifications of form are obtained at all is the point of interest here. This serves to strengthen the view which we present below, namely, that a temporary functional disturbance at earlier and “critical stages” (Stockard) may occasionally lead to the complete and the nearly complete suppression of gonad development. In other words, some abnormalities of form in gonads, as well as in other organs (Child, Stockard, Bellamy), may result from changes in metabolic rate—and, as I have shown (1923) in these same dove embryos, the effective change may be either an acceleration or a depression. In all this part of the field of teratogeny treated embryos of quite normal zygotic constitution develop some very abnormal structures, and sometimes wholly fail to develop other structures. These very profound developmental changes plainly rest on a physiological basis. The effort here is not to point to the specific thing that causes gonad absence, but to indicate the type of agency which results in that condition.

Possible Relation of Disease to Gonad Absence

We have indeed encountered a very few cases of the destruction of a gonad by disease ; the conditions attending these instances, together with the possibility of still other cases being effected by disease, must now be briefly considered. Tuberculosis is the only disease known to be or suspected of being in any way associated with the destruction of gonads in our material. In three cases, however, we are fully assured that such complete or practically complete destruction did occur.

In Case 26 the short descriptive account already given sufficiently reveals the conditions found and suggests the kind of evidence which establishes the destruction. The bird had laid many eggs, and was killed nine months after laying her last egg. The number of “single” eggs and thin shelled eggs laid by this bird is evidence that she was carrying some tuberculosis even while producing eggs. The presence of a large cheesy mass involving the left gonad region and spleen would have made us suspect the involvement of this gonad, even if we had not been in possession of data conclusively proving the earlier presence of an ovary.

In Case 34 conditions were essentially similar to those just described above. Case 34 showed the left gonad region involved (again with the spleen) in a large cheesy tubercle practically devoid of tissue. If a left testicle ever existed it was here destroyed by tuberculosis. The right testis had been pulled (probably also pressed upon) toward the middle line of the body by the growth in the left gonad region. The right testis of course may have supplied the sperm for the fertilisations known to have been effected by this bird. It is entirely probable, however, that a left testis was completely destroyed. The extremely small size (0.004 gm.) the right testis at death is accounted for by the universally observed reduction of testis size in pigeons dead of tuberculosis. (Riddle, 1918, Table IV., shows reduction in pure species to as little as 0.007 gm per testis.) It may be noted that in such cases, despite the great reduction of testis size, this organ does not lose its characteristic form and definitely testicular appearance. In the ovary of a tuberculous dove a similar reduction of the stroma (oöcytes larger than 1.0 mm. all atrophy) has never been observed, except when the ovary itself was the seat of active invasion and destruction.

Table IV.

Data for Twelve Cases of Extraordinary Size-Difference in the Two Members of a Pair of Testes.

Data for Twelve Cases of Extraordinary Size-Difference in the Two Members of a Pair of Testes.
Data for Twelve Cases of Extraordinary Size-Difference in the Two Members of a Pair of Testes.

The only other of the forty-eight cases here described which need be at all considered in this connection are Cases 19, 40, and 25. In the latter case a single right gonad was possibly destroyed by tuberculosis ; certainly a large tumour, not at all including the right suprarenal was found here. It would seem necessary to consider either a gonad or the Wolffian body as the seat of this tumour, which presented itself as. little more than a cheesy mass. The bird in Case 19 was killed by us, not by tuberculosis, though this disease was in evidence in spleen and liver. The left gonad—the only one present—was in no way directly involved, and the tuberculosis cannot be considered responsible for the “mere trace” of gonad present. A healthy member of its fraternity (Case 11) with “no gonads,” and another (Case 18) with a “mere trace of embryonic ovary,” provide the real clue to the conditions found in Case 19. The right gonad of Case 40 was only partially destroyed by a tuberculous (?) tumour, and of course similar tumours have been found in some other, birds not considered in this paper.

It remains to note that the destruction of the ovary by tuberculosis may be accompanied by a replacement of this organ by a testis. One case has been found in our studies and has been separately described (1924). An apparently identical case in the fowl was found and previously reported by Crew (1923). In both of these cases it is notable that it was, in all probability, only the left ovary that was destroyed by tuberculosis ; and that in the two cases described here (Cases 26 and 34) it was also the left gonad that was destroyed and by this same disease. Further, that in these two cases, precisely as in my case of sex reversal in the dove, the destroyed organs were involved with the spleen in a common tuberculous mass. Finally, it is to be noted that our study (1921) of the incidence of tuberculosis in the various organs of the pigeon has shown that in ring doves the spleen is the organ most frequently and most completely involved in this disease.

That the above type of substitution or replacement also occurs in frogs—in the reverse direction, testis replaced by ovary—has been made entirely probable in other studies by Crew (1921, 1921,a). In the newt, Champy (1921) has apparently further been able experimentally to effect the reduction of the testis followed by replacement with ovarian tissue. These studies of Crew and of Champy are of great interest to sex theory, and in so far as their results are concerned with the destruction of gonads they touch the point discussed here.

The above considerations perhaps sufficiently indicate the very restricted number of cases of partial or complete gonad absence chargeable to tuberculosis. Cases 26 and 34 are unquestionably different and distinct from their associated cases ; Case 25 is either similar to Cases 26 and 34 or it was a true case of no gonads. The two former cases are listed in this presentation not because there was doubt as to their distinction, but in order that every case in any way corresponding to our title headings might be presented. They are given places in Tables I. and II. for the same reason, but they there deserve elimination. The high incidence of these gonad deficiencies in “wide crosses,” and indeed in “first” and in “second” crosses constitute fairly crucial evidence that this type of circumstance largely conditions their appearance. Their frequency within single fraternities is evidence of the same kind. Many cases occur in entirely healthy individuals ; and finally, they occurred at all ages, as the following summary will show : A 14-day embryo ; one while hatching ; two at 2 days after hatching; one at 11 days after hatching (0.9 month); 1.3 months; 2.0 months; 3.5 months; 4.0 months; 10.9 months; 13.0 months; 14.3 months; 26.7 months; 50.3. months ; 183 months ; 215 months (three others—Cases 17, 25, 31—possibly entirely without gonads, were aged 6.4 months, 7.0 months, and 39.8 months).

All of the kinds of available evidence on this subject indicates that in these cases we are dealing with the absolute failure (Group 1) of the gonads to arise during development ; with an almost complete failure (Group 2) to develop ; or finally, with the complete failure of the gonad of one side (Group 3) accompanied by slight or by complete and normal development in the other. An attempt to interpret these conditions will follow in a succeeding section.

The Sex Behaviour of Some Special Cases

Reference has already been made in the condensed description of cases to the nature of the sex behaviour shown by many of these gonadless birds. It seems desirable to give a more detailed account of the behaviour of a few of these individuals. It is not the purpose here to consider at length the probable zygotic constitution, nor the special sex-determining influences, represented in each of the sixteen cases. That topic is being treated in a more appropriate place in an account of our studies now in preparation. The descriptions given here are intended to supply merely the fact of a demonstrated type of sexual behaviour in particular cases of gonad absence. Five cases will be more fully described, and but little discussion offered at this point.

Case 7.—The curve indicating the body size and the size changes of this bird is known from late 1912 to January 1917 (8th to 13th year). Its average weight (182 gms.) was notably less than that known (from about thirty weighings each) for three brothers (197, 215, 216 gms.). No sisters were alive at the period these weights were taken. In all similar crosses, however, it has been found that the normal male hybrids show higher body weights than do the females. Unquestionably the size of this gonadless bird was more nearly that of a female.

The behaviour record follows ; Attempts to mate this bird during its first five years of life were apparently not made (by Whitman). When nearly six years old it was placed with a brother (the sex of both being then unknown) and the two remained together during eight months. They refused to mate, both showing only male behaviour, and were separated when it seemed certain that both were males. Both were then placed in a pen of similar males (Fj generic hybrids) where only male behaviour was ever noted ; here Case 7 successfully defended itself and actively sought a female mate. When 13 years old it was taken from this pen and for two months (24th May to 20th July) was offered a female ring dove hybrid as a mate. No eggs resulted from this association ; and, though neither copulations nor any note concerning the cause of failure was recorded, the lack of egg-production by the female was doubtless the cause of the removal of the supposed male and the substitution of another. Our later records show clearly enough, however, that the failure of egg-production was not necessarily a failure of Case 7 to supply the necessary male or mating stimulus which the female requires for the production of eggs. The final records show that with the male next substituted she also persistently failed to yield a single egg before she was sacrificed (and found quite healthy) at the end of seven months. During her life this female had produced a total of one clutch, or two eggs—and these only at a very advanced period in life (13.1 months) for a bird of her race. Case 7 was again placed in the pen of male hybrids, and no doubt of its maleness had been raised during our period of more than ten years of observation. It fought until the end for a mate among these males and, though nearly 18 years old and carrying obvious tuberculosis in the left lung, its death was probably chiefly the result of the last fight or fights of a career of fighting for a mate. The two thyroids were markedly different in size, the right weighing 0.0052 gm., and the left, 0.0100 gm.

The weight of this bird was more nearly that of a female; ten years of behaviour was that of a male ; the only reproductive organ present was the dwarfed left oviduct of the female.

Case 12.—The weight curve is known from the autumn of 1912 to January 1917. This curve is practically indistinguishable from that of three brothers, except that its weight throughout is somewhat less (198 gms. to 205, 203, 217 gms.) than in either brother. Four sisters, for the same period, gave average weights of 191, 188, 197, 192 gms. Its body weight, therefore, was intermediate to that of normal males and females.

A fuller description of the masculine sex behaviour of this bird follows: When 5 to 8 years old, and while mated to a sister, its complete and unqualified maleness of behaviour was fully established (a previous mating as a male with another sister has been noted in the condensed description given above). During nearly three years it continuously stimulated the production of eggs (copulating as male) in its hybrid mate; during eleven months of one year (1913) the large total of 72 eggs were thus laid by its mate. At one or another time (exact dates all known) eggs of other birds were given to this pair, and Case 12 then perfectly incubated eggs and young at the periods normal to the male. None of the eggs produced by either of the mates of this bird was tested for fertility (used in studies on yolk size), since “tests” would have little chance to reveal anything—the inter se crosses of male and female Fj hybrids of this type usually proving completely sterile. The last five years of this bird’s life were spent with a group of males only (F1 generic hybrids) and here it—like all its male companions—fought till the end for a mate. When finally killed for examination these fights has left it without a feather on head, neck, and extreme anterior parts of body. The two thyroids of this bird were of quite unlike size; the right, 0.0096 gm. ; the left, 0.0168 gm.

The space between the pubic bones of this bird (at 7 years) was 9.5 mm. In five brothers this space averaged 9.0 mm. (6.5 to 10.5); and was 11.4 mm. in seven sisters (9.0 to 14.0).

It appears therefore : (a) That the pubic bones of this gonadless bird suggest maleness rather than femaleness. (b) That the sex behaviour of the bird, during eleven years of observations, was normally and characteristically that of a male; even the recurrent instinctive behaviour which is dependent upon the various stages of the reproductive cycle were normally masculine, (c) That the size (indicated by about thirty weighings) of the bird was slightly larger than its largest sister, and smaller than its smallest brother—i.e. it was a size intermediate of the two sexes, (d) That when killed for examination, practically healthy, it possessed neither trace of gonads nor discoverable trace of any sexual ducts whatever.

Case 15.—The size of this bird was notably less (128 gms.) than that of even the females of this species (about 165 gms.). This abnormally small size also complicates the mating record of this bird, since aggressive masculine behaviour is not freely expressed toward a much larger bird. Definite details of the behaviour of this bird are also lacking. Its life was relatively short—the period of its maturity lasting only a little more than a year, and at least the latter part of this was complicated by the presence of disease. It is definitely known only that during the period three to eight months before its death it failed to secure as a mate a mature female of its own kind ; that this female, however, was of a weak and short-lived race and that her death from disease occurred at the close of the attempted mating. Also that Case 15 further failed—during two months—to win a female ring dove (150 gms.) as a mate. These few facts, however, make it entirely probable that this bird was neither markedly masculine nor feminine in its behaviour. Two normal females will usually mate with each other and both produce eggs within such periods.

In this bird the body size constitutes a definite abnormality which is not distinctive or indicative of sex. The behaviour is not certainly known, but it was probably approximately neutral. The left oviduct—despite the progress of disease sufficient to cause death—was better developed than has hitherto been observed either in cases of “no gonads,” or in the considerable number of cases of oviducts associated with testes. Altogether, the evidence here indicates that ovaries rather than testes would have fitted more properly into the other characteristics of this individual.

Cases 18 and 19 are of somewhat less significance than the preceding cases, because a “mere trace” of one gonad was found in each, and the behaviour data are very incomplete for Case 18. The further fact that Case 18 was dead of tuberculosis might also suggest to some that the “trace” found was perhaps but a remnant of an earlier gonad reduced by this disease. When all of the facts concerning these two cases are set forth, however, they each acquire some importance. Since both cases arose within the same fraternity a part of the following statement applies to both, and a further special statement on Case 19 is perhaps not required.

Case 18 had a “trace of immature left ovary” when dead of tuberculosis at 28.5 months. Some hundreds of similar cases of tuberculosis in much younger birds have failed to show any tendency to cause either the type of immaturity shown in this gonad, or to cause atrophy of ovarian tissue when this tissue was not itself the seat of disease (Riddle, 1916). Moreover, one other of the thirteen sisters of this bird, killed healthy at 4.0 months, also had a notably immature and quite small (though not extremely minute as in the present case) ovary. A fuller statement concerning the prevalence of gonad abnormalities in the entire fraternity to which Cases 18 and 19 (and 11) belong will further confirm the above view ; and, in addition, such data will afford further specific evidence that the types of gonad deficiency dealt with in this paper are developmental, not disease phenomena, and are closely connected with hybridity. Of 31 pairs of gonads examined in this fraternity, 10 were normal testes ; 3 were abnormal or atypical testes ; 7 were normal ovaries ; 7 were abnormal ovaries; in 4 others (including Case 11) macroscopic examination of the gonads failed to designate the sex of the individuals.

Case 18 had wide pubic bones and was first offered a male of her kind as a mate. It is not known whether she either accepted or copulated with this male; but she laid no eggs and never “nested without eggs.” It is certain, therefore, that she either failed during about ten months to mate with an active male; or, having mated, the normal stimulus of copulation here failed to induce either ovulation or nesting behaviour in this bird. Following this long trial as a female this bird was given a female mate. The latter female was laying when Case 18 was substituted for her male mate; she did not lay again. In this way we obtain some indirect evidence that Case 18 did not here function as a male. To us it is practically certain—both from the recorded data and from the very absence of certain other data from her record—that this bird with a “mere trace of immature ovary” showed nothing resembling the aggressive and continuous fight for a female mate which characterised some of the birds entirely without trace of gonad earlier described; and also nothing comparable to the behaviour of Case 19 (her brother or sister), in which also a “mere trace of gonad of unknown nature” was present

It is convenient to discuss only one point here. We are further compelled to consider that the two cases (18 and 19) last described show different or divergent sex behaviour ; and that the “trace” of gonadal tissue present can hardly be presumed to have either initiated or to have notably affected this divergence (a similar divergence occurred in the total absence of such tissue in Cases 7 and 12 on the one hand and in 15 on the other). In these cases we conceive the same conditions (zygotic ; direct effects of wide hybridisation ; reproductive work, etc.—all probably summarisable in the term “metabolic rate or level”) which decide the type of sex differentiation, including gonad formation, as also being effective in establishing divergence of behaviour even in the absence (complete or nearly) of gonad formation. If accessory structures (gonads) which normally aid or emphasise the expression of sex are not formed—if these functions are not developed within a highly specialised organ—the remainder of the organism retains and expresses the type of sex function which was impressed upon it and maintained, within it A quite different picture is presented in castration experiments ; for, these are cases in which at least certain aspects of this function during development has been actually released or relegated to a specialised region—and this region is then bodily and brusquely extirpated.

Further discussion of the rôle of the gonad incretions will follow a consideration of an aspect of the origin of the gonads.

Some years ago (1916) I reported data indicating that in healthy pigeons the right testis is usually of larger size than the left Even at that time we were inclined to suspect that this particular size-difference of the testes, together with the prevailing atrophy of the right ovary of birds, was in some way related to results then recently announced by Firket (1914) and by Swift (1914, 1915). These workers had independently shown in 1914 that the “primordial germ cells” of the fowl arise in a crescent shaped area (extra embryonic) distinctly anterior to the head process ; that these cells then become distributed to all parts of the embryo. It was further observed that between the 20- and 25-somite stage these cells collect or concentrate in the splanchnic mesoderm and blood vessels, particularly in and near the region of the right and left germinal ridges. Here their later development became of course an intimate part of the development of the gonad—ovary or testis. Still another important observation made by these workers was noted in our 1916 publication (p. 87) in the following terms :—

Firket (1914) and Swift (1915) have shown that in the chick embryo there were more primordial germ cells in the left gonad, and that this gonad is there also distinctly larger than the right Allen (1907) found that the sex cells were unequally distributed to the two gonads of the turtle, the left receiving most. In this form only 24 to 70 per cent of the sex cells ever enter the gonads…The meaning of this pronounced inequality in the distribution of the primordial germ cells, which is plainly associated (at a very early stage) with a larger left embryonic gonad … is by no means clear. But, whatever this may mean, it is probably a situation of importance to the theory of sex.

Firket and Swift agree that two to five times as many primordial cells collect in the left germinal region as in the right. Let us next note that also in the turtle Allen found the left gonad anlage to be the larger ; and that, certainly in this case, a notably high percentage of the sex cells entirely fail to enter the anlage of the gonad. Again, in the fowl, Swift (1914) noted that in “one abnormal case four masses of germ cells were in the blood vessels of the head” in an embryo of 25 somites. The point of interest here is that this latter type of abnormality should occur at all ; and it would seem from Swift’s data that this degree of failure of these cells all to reach the site of the future sex glands occurred once in about fifteen cases in which it could have been observed. It may be recalled that the migration of the germ cells, to one or another extent, has been very generally observed in higher animals.

We now ask, What would be the development history of the genital ridge region if this area were completely deprived of the primordial cells (the Keimbahn) ? That a Wolffian body might form would perhaps be thought probable or conceded by all. It is our own general experience, however, that if it did form in our “no gonad” cases its resorption was usually complete ; and the occasional absence of sexual ducts possibly renders its temporary existence a matter of doubt. That a true gonad would form and function under the conditions postulated is a proposition challenged by very much of our present knowledge of differentiation.

Without elaborating an argument on a subject which now cannot be definitely resolved, we may state the position taken here as follows : (1) The known facts make it probable that very rarely all primordial cells fail to reach the site of the gonad. The relatively limited number of such primordial cells ; their favoured accumulation in greater number on one of the two sides of the embryo in both of the two Sauropidans studied ; the normal failure of many or most of these cells to enter the gonad of one of these two forms—and a single, abnormal, otherwise probably similar case in the other ; the origin of these cells in birds in distant extra-embryonic regions, from which they reach their distination only through the hazards of the early embryonic circulation ; their scattered position throughout the embyro at an intermediate stage—all this suggests to us that it sometimes happens that no primordial cell reaches either genital ridge (conceivably also conditions similar to those which produce other experimentally induced teratogenies may occasionally prevent the formation, or even the usual expression of the action, of these cells). (2) When no primordial sex cells enter the germinal ridge no gonad is formed. It may be pointed out that this in no way precludes the possibility of instances of the formation of germ cells from peritoneal cells ; since, if this does in fact occur, the process may be conditioned by the circumstance that the “Keimbahn” cells were, or at some time had been, in contact with them, and thus influenced their capacity to form germ cells in much the same way as contact of the optic cup with ectodermal cells conditions the power of the latter to produce lens cells.

The known facts therefore warrant the assumption that in rare instances primordial cells may fail to reach both the right and left gonad regions. Our sixteen (probably eighteen—Nos. 17, 31) cases of this sort would require this dual absence to occur once, in the formation of something more than 1000 embryos. Similarly, it it conceivable that a single primordial cell is insufficient to produce a complete normally functioning gland (not an entirely satisfactory assumption). There are approximately nineteen (or twenty-five) known cases (plus perhaps ten not detectable in females — see below) to be accounted for in our data ; here also, therefore, this insufficiency is demanded about twice in 1100 embryos.

Again, if this principle is at work at all it should further occasionally chance that no primordial cells reach the future site of one gonad but do attain the other. Our data include fifteen such cases from Group 3, to which four or five cases (19, 20, 25, 30, and ? 29) must be added from Group 2 ; perhaps six (or twelve) additional cases should be added from Table IV.—though more than “traces” of testis were present in those cases. Finally, the absence of these cells from the right gonads of post-embryonic females could never be detected, and this would add perhaps ten cases to the above twenty-six. The required frequency of this occurrence is therefore essentially that of the preceding type of deficiency. Thus about 107 gonads in approximately 20,000 pairs—or one of each gonads—are conceived as having been completely deprived of, or restrictively provided with, these primordial germ cells (mostly in hybrid offspring).

It was stated immediately above that approximately nineteen instances of minute non-functional traces of gonads have occurred. In fixing this number it is necessary to realise that traces of right ovaries in birds after hatching are not abnormal and cannot be detected ; and also that a few cases offer special difficulties of classification. More specifically we find that the right gonad provides seven cases (23, 28, 29, 32, plus three cases to be described elsewhere in a discussion of hermaphroditism); the left gonad twelve cases (18, 19, 21, 22, 23, 24, 27, 29, 30, 32, 33, 38). The circumstance that the abnormal traces of right ovaries cannot be distinguished in post-embryonic females, while absent left ovaries are recognisable at all stages, would doubtless more than offset the apparent excess of left gonads thus affected (seven of the above twelve were recognisable as ovaries, none as testes). Six additional pairs of testes (for adult cases, probably approximating to these non-functional gonads) should perhaps be added to this list as given above ; these consist of four extremely diminutive adult rights and two lefts (Table IV.). Those six cases, however, will be specially considered below in connection with adult testis pairs showing extraordinary disparity of size.

Now it happens that a partial test of the truth of the conception just outlined above is theoretically possible through an analysis of the distribution of all the observed gonad deficiencies. In the case of pairs of testes in adult males a very effective test would be provided by a sufficient number of the deficiencies. In this sex both gonads develop and continue in evidence. In adults the question of a mere retardation or delay in development of the diminutive gland can scarcely be conceived. The other element necessary for such a test is provided in the circumstance that two to five times as many primordial cells normally enter the left gonad as enter the right (fowls). The latter gonad, therefore, should fail to develop—if failure results from this cause—decidedly more often than does the left. Our own available data of this kind are summarised in Table IV. If the very heavy restrictions placed upon the development of one of a pair of testes is of the nature suggested, we shall expect to find but few cases (in adults of one sex only) even in the examination of some thousands of testis pairs. We have here taken a size disparity of two and one-half times as indicating a real initial disparity in the developmental impulse (six of these twelve testes of reduced size were doubtless functional). On this basis we find nine cases of deficient right testes and three deficient lefts (3:1). If a smaller size difference (1.5 or 2.0:1 instead of 2.5:1) be made the basis of comparison the ratio of deficient rights is greater than 3:1. If this basis be so made that one gonad must be five times the size of the other then a ratio of 5 : 2 is obtained (see Table IV.). From these comparisons, therefore, it is learned that these extraordinarily small testes of adults have actually been more frequently encountered than diminutive lefts.

For a second and approximately similar test we may compare the numbers of absent or almost absent right and left gonads of the males (and those apparently male) included in Groups 2 and 3. Here we shall find six cases (19, 37, 41, 42, 44, 48) of such defective rights ; and only two cases (43, 47) of defective lefts. A third test, complicated by the fact that neither absent nor very minute right ovaries are detectable in post-embryonic females, is given in the second paragraph above ; only seven rights to twelve lefts are found, but the data are subject to the probabilities noted” there. A fourth test is afforded by the cases showing complete abnormal absence of one gonad. Here there are nine or eleven absent rights (22, 30, 35, 36, 37, 39, 41, 42, 45, and ?i9, 38) ; and five absent lefts (25, 40, 43, 46, 47). These four partial tests are supplied by our data and they indicate with fair consistency that the right testis is absent—and perhaps also nearly absent—two or three times more often than is the left. The four cases earlier cited of the “one gonad” condition in other vertebrates show three rights absent and one left absent.

In connection with the point under examination, it remains to note that for the right testis in pigeons to be even slightly smaller than the left is a violation of a widely established normal order (1916, 1918). Further, it has recently been made clear (1924 a) that the right gonad region in pigeons, and in at least some other birds, is more favourable than is the left for the growth of testicular tissue ; and conversely the left is the more favourable for ovarian development. In the cases listed in Table IV., therefore, the deficient testis rudiments located on the right side presumably had some influence helpful to the attainment of increased size, which would not be exercised in favour of testis rudiments developing on the left side. It is thus made the more remarkable that utter failures, or nearly complete failures, of the right gonad have proved more numerous than failures of the left. This becomes not only intelligible but necessary if no basis whatever for gonad development is most frequent on the right side.

The facts now available in the several sections of this paper will enable the. student who has not had the opportunity to inspect a bird with “no gonads” to realise the hopelessness of obtaining anything for sectioning from these cases. Having early noted the result of Poll’s effort, and in each case being visually confronted with the futilities involved, we have never attempted to have sections made of any portion of one of these birds. It can scarcely be contended that traces were ever formed, and they probably had not a chance to form, in our sixteen cases. In the complete absence of the gonad it will likewise scarcely be contended that any true interstitial tissue could be formed elsewhere in the body, since the differentiation of this type of tissue is probably conditioned by its association in a particular type of organ.

Most cases of gonad absence hitherto encountered in birds are clearly the expression of a developmental error and are not the result of disease. The analysis given of these cases makes it seem entirely probable that the same thing holds true for most of the cases observed in other animals.

The nature of the primary cause of gonad absence is at best a difficult subject. Since this point is here attacked apparently for the first time it is doubtless treated with very partial success. The subject too is many-sided. Whitman’s (1919, p. 90) data permitted us to write (in 1915) that “some hybrids—a much higher proportion than in pure breds—develop no sex glands, or they develop diminutive or otherwise abnormal ones.” In the present paper, and in studies of others published since that was written (and cited in our introductory remarks), the full truth of that statement has been made clear. It is now entirely evident that hybridisation is associated with these cases in unexpected frequency.

A next question easily arises, Does the absence of a gonad imply some abnormality of chromosome distribution — so frequent and familiar in hybrids—which prevents or interrupts some pre-gonad stage of the formation or division of the cells of the “Keimbahn,” while other cells continue much less disturbed ? Or, is this developmental failure to be placed on some more immediate and physiological basis ? We have wholly inclined to the latter view, and have attempted a constructive account of a series of facts which to us seem to afford it a basis ; but we can scarcely hope that the facts proceeding from our own experience and material will impress others as much as they influence us. A part of the importance of cases of “no gonads” rests upon the circumstance that, in adequately known material, they seem capable of contributing to sex theory from a number of unexpected sources.

Some data concerning the sexual ducts have been presented in connection with our case descriptions, but a consideration of the significance of these sexual characters has been undertaken in another more inclusive, if not more suitable topic. The data of this sort given here, however, would seem to be sufficient to demonstrate that in birds it is quite possible for the oviduct to attain a fair degree of development wholly apart from the presence of an ovary.

A very special importance attaches to the sex behaviour of a gonadless vertebrate. We shall here recall two points demonstrated by our experience as recorded above. (1) All birds of this kind do not show essentially similar behaviour. It is true, however, that very few of them demonstrated feminine or neuter behaviour. Elsewhere our account of hermaphroditic doves will again show the same thing. (2) Some doves without trace of gonads—and one without even a discoverable trace of sexual ducts — have demonstrated, continuously and over long periods of time, the most accentuated masculine behaviour. We have also noted the degree to which other masculine so-called “secondary” sex characters were developed in these particular birds.

The repeated observation of exclusive and emphatic masculine behaviour in this type of bird seems to make it certain that these expressions of maleness do not here rest upon an internal secretion of the testis—neither of interstitial nor of reproductive testis. Confronted with this fact is it not necessary to recast our expressions, perhaps some of our ideas, concerning the formative or directive properties of gonad incretions ? No lengthy consideration of this specific point may be undertaken here. To avoid being quite misunderstood, however, it may be said that I myself have early reported (1914) briefly a fact showing the unlike effects of testis and ovarian extracts and suspensions on the sex behaviour of pigeons. There is no attempt made here to question the frequently observed developmental differential—the specificity of the influence on other organs or functions—involved in something that is associated with testis on the one hand and ovary on the other. But, it is suggested that the same basic thing or condition which decides the type of gonad—testis or ovary — decides also and at the same time the type of interrelationships this gonad shall set up or facilitate in the organism. These interrelationships, moreover, must usually and normally be neither antagonistic to the type of gonad nor to the type of basic condition which determines the gonad ; if the principles of development proceeded otherwise most adult animals would be not organisations, but rather congeries of incompatibilities. If therefore the gonad incretion is essentially the agent or amplifier of the condition which determined the type of gonad it is possible to understand how, in the complete absence of both the gonad and its incretion, the specific type of development or reaction may still be effected.

At the present time many investigators in this field seem to regard the internal secretions of the gonads as the directors or controllers of the sex type ; and these incretions are themselves not regarded as controlled by anything. Or, if controlled, this is conceived as being accomplished by the zygotic endowment of the animal. Reciprocally, those who regard sex determination and sex differentiation as purely and inevitably a matter of zygotic endowment will concede that a part of the mode of action of the sex factors consists in the development and utilisation of the specific incretion of each gonad. On the other hand, many students of hermaphroditism emphasise the lack of sharp divisions in anything concerned in sex, and sometimes report conditions confounding alike to the geneticist and incretionist. Again, experimentalists working with sex have delivered into our hands individuals whose sex is the opposite of their cytologic “determination,” and whose incretions also are first established and then reversed or defeated. These changes and reversals follow from a variety of procedures.

In these separate aspects of the study of sex have we not been dealing with conceptions which are true but only partly true ? Concerning the gonadal internal secretions, as the subject immediately in hand, we here emphasise the view that they are not only controllers, but they are or may be themselves controlled (and the same may be said of the genetic or chromosome basis of sex). In the case of sex both the type of incretion and the genetic factor respond to a more basic influence or condition. We know no fact in opposition to this statement. All the facts of hermaphroditism readily fit themselves to this conception. The facts dealing with the experimental transformation or modification of sex fit no other conception. But do the facts from this latter field, obtained on quite varied material and with methods apparently widely different, lead to any common view as to what this “basic condition” is ? Quite clearly it is to the results in this latter field that we must look for the answer—which, if wholly true, will doubtless be a unanimous one. Within the last several years the writer has contributed some bits of evidence that this primary and fundamental basis and distinction of sex is a differential in metabolic rate, or metabolic level. In the section of our study presented here, and in the other aspects of the study of which this is a part, we find some further confirmation of that conclusion.

7.
Adlerz
(
1908
), “
Zwei Gynandromorphen von Anergates atratulus Schenck
,”
Arch, für Zool.
,
5
.
Allen
(
1907
),
“A Statistical Study of the Sex Cells in Chrysemys marginata,”
Anal. Rec., 1 (also Anat Anz., 90)
.
Bond
(
1914
), “
On a Case of Unilateral Development of Secondary Male Characters in a Pheasant, with Remarks on the Influence of Hormones in the Production of Secondary Sex Characters
,”
Journ. of Gen.
,
3
.
Champy
(
1921
),
“Changement experimental du sexe chez le Triton alpestris Laur
.,
Compt. rend. Acad, des Sci.
,
172
.
Cockayne
(
1916
),
Trans. Ent. Soc.
,
London
.
Crew
(
1921
) “
Sex-Reversal in Frogs and Toads. A Review of the Recorded Cases of Abnormality of the Reproductive System and an Account of a Breeding Experiment
,”
Journ. of Gen.
,
11
.
Crew
(
1921a
), “
A Description of Certain Abnormalities of the Reproductive System found in Frogs, and a Suggestion as to their Possible Significance
,”
Proc. Roy. Phys. Soc.
,
20
.
Crew
(
1923
), “
Studies in Intersexuality, Part II.: Sex-Reversal in the Fowl
,”
Proc. Roy. Soc. B.
,
96
.
Firket
(
1914
), “
Recherches sur 1’organogenese des Glandes Sexuelles chez les Oiseaux
,”
Arch, de Biol.
,
29
. (Also Arch, de Biol., 1919-1920, 30.)
Goldschmidt
(
1916
), “
Experimental Intersexuality and the Sex Problem
,”
Amer. Nat., 50
.
Harrison
(
1920
), “
Genetical Studies in the Moths of the Geometrid Genus Oporabia (Oporinia), with a Special Consideration of Melanism in the Lepidoptera
,”
Journ. of Gen.
,
9
.
Hegar
(
1878
),
In Sammlung klin. Vorträge, by R. Volkmann, Nr
.
136
138
.
Jacquet
(
1895
),
Bibl. Anal
,
8
,
Paris
.
Loisel
(
1901
), “
Grenouille femelle présentant les caractères sexuels secondaire du male
,”
Compt. rend. Soc. d. Biol.
Morgan and Bridges
(
1919
),
The Origin of Gynandromorphs, Carnegie Inst Wash., Publ. No. 278
.
Nagel
(
1897
), “
Entwicklung und Entwicklungsfehler der weiblichen Genitalien
(In Veits Handb. d. Gynäkol., 1, Auflage 1)
.
Poll
(
1909
), “Zur Lehre von den Sekundären Sexualcharakteren,”
Sitzungsber. d. Ges. Naturforsch.
,
Freunde zu Berlin
.
Pütz
(
1889), Quoted from Sauerbeck
(
Frank. Zeitschr. f. Path.
,
8
, 1909).
Riddle
(
1914
), “
A Quantitative Basis of Sex as indicated by the Sex Behaviour of Doves from a Sex Controlled Series
,”
Science, N.S
.,
89
,
440
.
Riddle
(
1916
), “
Size and Length Relations of the Right and Left Testes of Pigeons in Health and Disease
,”
Anat. Rec.
,
2
.
Riddle
(
1918
), “
Further Observations on the Relative Size and Form of the Right and Left Testes of Pigeons in Health and Disease, and as influenced by Hybridity
,”
Anal, Rec.
,
14
.
Riddle
(
1920
), “
Differential Survival of Male and Female Dove Embryos in Increased and Decreased Pressures of Oxygen : A Test of the Metabolic Theory of Sex
,”
Proc. Soc. Exper. Biol, and Med.
,
17
.
Riddle
(
1921
), “
Incidence of Tuberculosis in the Various Organs of the Pigeon
,”
Journ. Infec. Dis.
,
29
Riddle
(
1923
), “
On the Cause of Twinning and Abnormal Development in Birds
,”
Amer. Journ. Anat.
,
82
.
Riddle
(
1924
), “
A Case of Complete Sex-Reversal in the Adult Pigeon
,”
Amer. Nat.
,
58
.
Riddle
(
1934a
), “
Sex in the Right and Left Sides of the Bird’s Body
,”
Proc. Amer.. Phil. Soc.
,
68
.
Smith and Haig-Thomas
(
1913-14
), “
On Sterile and Hybrid Pheasants
,”
Journ. Gen.,
3
.
Swift
(
1914
),
“Origin and Early History of the Primordial Germ Cells in the Chick,”‘
Amer. Journ. Anat.,
15
.
Swift
(
1915
), “
Origin of the Definitive Sex-Cells in the Female Chick and their Relation to the Primordial Germ Cells
,”
Amer. Journ. Anat.
,
18
.
Whitman
(
1919
),
Posthumous Works, 2, Carnegie Inst of Wash., Publ. No. 257
.
*

It is elsewhere shown that this condition does not fully fix the sexual status among pigeons, since we have found several cases of a left oviduct associated with two testes (normal, but of reduced size) in these forms.

*

It should perhaps be noted that these different treatments were not equally severe as measured by the percentage mortality of the treated embryos (the actual treatment of testis-bearing and ovary-bearing embryos within each of the three types of treatment was of course the same). In “increased oxygen,” 40.2 percent, were killed (or dead within forty-eight hours) ; in “decreased oxygen,” 44.0 per cent ; in low temperatures, 48.2 per cent