ABSTRACT
The experiments of Guyer and Smith have attracted wide-spread attention. Their results are recorded in two papers (1918 and 1920). The first paper describes the injection of rabbit lens into fowls and the subsequent injection of the sensitised fowl serum so produced into pregnant does. Some of the young of the treated does exhibited certain abnormalities of the eye. No breeding experiments are recorded in this paper. In the second paper a similar series of experiments is described, as a result of which more abnormalities were produced. Breeding experiments with the abnormal young showed that the defects were inherited. In view of the interest which these experiments have aroused, it seemed desirable to repeat them, and this paper records an attempt to do so. The methods used by us were modelled on those employed by Guyer and Smith ; the particulars in which our methods differed from those of these two authors will be indicated in the description of our experiments and attention again drawn to them, so far as the matter seems relevant, in the comparison of our results with those of Guyer and Smith in the discussion which will follow.
1. Injections into Fowls
a. Methods
The fowls used were mongrels. They were weighed regularly during the period when injections were being given. We aimed at giving injections about every third day, but when a fowl under treatment exhibited any considerable loss of weight, further injections were postponed until weight was regained. Fowls Nos. I. to V. were two or more years old and frequently exhibited loss of weight, as a result of which the intervals between the injections were often longer than desirable. The remaining fowls were about a year old and little trouble from loss of weight was experienced when dealing with them. Except in the cases of shock to be described later, which, if it occurred at all, occurred immediately after injection, the younger fowls did not seem to suffer in health as a result of the treatment.
The lenses used for injection were removed from rabbits under sterile conditions immediately after the death of the animal. For certain reasons to be given later, ox lens was sometimes used in place of rabbit lens, but the methods employed were precisely the same as in the case of rabbit lens. The lenses were placed in a mortar and, after the addition of a few c.c. of normal saline solution, were thoroughly pulped. However thoroughly the pulping was done, there was always a certain amount of gelatinous residue which, if drawn into the syringe, was found to clog the needle. If the liquid was merely allowed to settle and the supernatant fluid drawn off, it was found that the fluid often still contained enough gelatinous substance to cause trouble, and therefore in most of the experiments the pulped fluid was centrifuged. The latter method proved quite satisfactory, but when towards the end of the experiments no positive results had been obtained, it was thought that criticisms might be made of our methods on the ground that the whole of the lens substance had not been injected. Centrifuging was therefore abandoned, and in the later experiments injections made with as much of the liquid as was possible without causing complete clogging of the needle.
In most of the experiments made by Guyer and Smith the intraperitoneal method of injection was used. In a few cases they employed the intravenous method. They state that the latter gave better results but that it was difficult to perform. That it should give better results is to be expected. It was therefore determined to try the intravenous method from the outset. For this purpose we used the wing vein, and after a little experience the operation was not found to be difficult. The vein, however, is small, and from time to time failures occurred, and under these circumstances we used the intraperitoneal method in order that there should not be too long an interval between the injections. Further, in the case of fowl No. VIII. only the intraperitoneal method was used so that the results of the two methods might be compared. The use of the intraperitoneal method presents no difficulty whatever. The fowl is laid on its back, a fold of the abdominal wall is held between two fingers and the needle inserted. When employing the intravenous method it was found best to remove a small piece of skin before inserting the needle. The removal of the skin sometimes caused bleeding, and the position of the vein was thus obscured. This was the principal cause of such failures as we had that were not due to the clogging of the needle owing to the presence of gelatinous residue in the syringe.
In Table I. are given details of all the injections into fowls performed by us. The table shows the date of each injection, whether rabbit or ox lens was used, the amount of fluid in the dose, the amount of lens in the dose, and the method employed, whether intravenous or intraperitoneal. The amount of lens is given in fractions of a lens in the earlier experiments and in grams in the later experiments. It was sometimes convenient to use the lenses of half-grown rabbits and, as under these circumstances fractions of a lens are misleading, the lenses were weighed and the dose recorded in grams. For the purpose of comparing the earlier with the later experiments the weight of a rabbit lens may be taken as 0.5 g. Doses of ox lens are always recorded in grams. The weight of an ox lens is about 2 g.
The last column of Table I. shows that in certain cases fowls suffered from shock after treatment. If shock was experienced the fowl either died, or else recovered completely within a very few minutes. The nature of the shock was not determined, and the pathologists whom we consulted were unable to give us any definite information. It was not anaphylactic in nature because the interval between the injections was too short — namely, on the average, about three days. 1t had nothing to do with centrifuging or not centrifuging the liquid because it occurred under both circumstances. It is also worthy of note that on more than one occasion a fowl suffered from shock after receiving an injection consisting of part of the liquid in the syringe, when other fowls, who received the rest of the liquid in the syringe, showed no signs of shock. But on two of these occasions the fowl to exhibit shock was the last fowl out of three to receive an injection from the same syringe, and the fowl to suffer may therefore have received more sediment in the dose than did the others. It seems that the thicker the liquid the more likely is shock to occur.
b. Results
Except in the cases of the first fowls employed precipitin tests were made from time to time, according to the method elaborated by Professor Dreyer, to ascertain whether, as a result of the treatment given, antibodies reacting with lens were present in the blood of the fowls. Blood was withdrawn from the fowls by puncturing the wing vein with a sharp scalpel and collecting the blood with a Pasteur pipette. In one case only was infection subsequently observed. The blood was incubated for two hours, the serum separated from the clot and then centrifuged. A series of tubes was then made up giving dilutions of serum of decreasing strength. In the first tube 15 drops of lens suspension were added to 10 drops of serum ; the lens suspension in this and all other tubes was of the same strength—namely, 39 parts of normal saline to 1 part of lens. In the second tube 15 drops of lens suspension were added to 5 drops of serum to which 5 drops of normal saline had been previously added. Thus the first tube gave a dilution of 0.4 of the original lens suspension and the second tube a dilution of 0.2. In this manner a series of tubes was made up—the last having a dilution of 0.002. Experience showed that reactions were never obtained in solutions of less than 0.04 dilution, and therefore only dilutions of 0.4, 0.2, 0.08, and 0.04 were subsequently employed. In all cases a control was made up of lens and normal saline alone. The tubes were then incubated for about two hours and afterwards examined at varying intervals. The amount of precipitation observed was recorded each time the tubes were examined. The method of recording the amount of precipitation was suggested by Professor Dreyer. Three terms were used : — (1) Total, indicating complete precipitation ; (2) standard, indicating a rather less degree ; and (3) trace and definite trace indicating a less but definite amount. These three grades correspond to the symbols “+ + +,” “+ +,” and “+, ” used by many serological workers. “Trace —” and “trace?” were also used, but not regarded as definitely positive. It was frequently observed that the reaction was delayed, and that the full reaction was not visible until twenty-four or even forty-eight hours after the removal of the tubes from the incubator.
Table II. summarises the results of these tests. For each test made there is recorded in the table :—the date; whether the test was against rabbit or ox lens, or both ; the minimum strength of dilution which exhibited a reaction reaching the grade “definite trace”; and the number of hours after removal from the incubator at which the reaction was first observed to reach its maximum. Thus the table shows that fowl No. IX. was tested on 6.12.22, and that after eighteen hours the reaction had reached its maximum, when a precipitation of the grade “definite trace” or “+” was visible in the tube containing a dilution of o.2 but was not visible in the lower dilutions. When the tube exhibiting a precipitation of the grade “+” is not that containing the highest concentration of serum, as, for instance, in the case of the test referred to above, it is almost always found that the tubes containing the stronger dilutions showed precipitations of a higher grade, e.g. “standard” or “total.” Early in the work, observations after more than twenty-four hours were not taken. The maximum reaction is usually at about forty-eight hours.
These tests give definite results and afford valuable evidence, but too much importance should not be attributed to their quantitative comparison. Variations in temperature and in other factors may increase or inhibit precipitation, and in this manner may be explained the fact that some fowls which showed a considerable reaction after a number of injections, showed no greater or even a slightly smaller reaction after a further series of injections. These considerations, however, do not in any way invalidate the following conclusions which may be drawn from the experiments. (a) Every injected fowl that was tested showed a positive reaction, with the exception of fowls Nos. IX. and XV. which, though they showed a well-marked reaction after the first series of injections, were negative to rabbit lens after a further series of injections. They were, however, positive to ox lens and, as the tests with rabbit lens were made the same day, it may be that this anomalous result is due to some defect in these particular experiments. (b) Five uninjected fowls were tested (Nos. XIX., XX., XXVIII., XXIX., and XXX.) and they were all negative. These fowls were then injected ; two died of shock before they could be tested, but the three others on being tested all gave positive results. There seems, therefore, no doubt that antibodies against lens can be produced by this method, and are not normally present in fowl’s blood. (c) The strength of the reaction in the animal decreases very slowly even after a considerable lapse of time. Thus fowl No. XI. gave a reaction “+” in dilution 0.4 on the sixth day after the last of a series of six injections. No doubt the full effect of the injections had not by that time been felt. It is nevertheless remarkable that more than three months later, during which time no further injections had been given, a reaction was observed on testing in the dilution of 0.08. (d) Almost as strong reactions against ox lens as against rabbit lens are obtained with the serum of a fowl treated with rabbit lens, and vice versâ. This indicates a high degree of organ-specificity as regards the lenses of different species of Mammals, as has been previously established by other workers (see Hektoen, 1922, for confirmation and extension of the early work of Uhlenhuth). This observation led in the later part of our experiments to the use of ox lens in place of rabbit lens. If the serum of a fowl treated with ox lens can precipitate rabbit lens, it seems to follow that such serum should when injected into rabbits be able to produce the same effects as the serum from a fowl treated with rabbit lens. Ox lens can be obtained fresh from the slaughterhouse for the asking, while a supply of rabbit lens involves considerable expense.
2. Injections of Serum into Rabbits
a. Methods
All the does used were of the common agouti type, with the exception of three which were of the Dutch breed. Two bucks were employed, one of which was a self-coloured blue (X. 2) and the other a Japanese (X. 1). It was not found necessary to kill a fowl in order to obtain blood each time an injection was to be made as was practised by Guyer and Smith. The method followed was to wash the ventral surface of the wing with carbolic soap and water and, after swabbing the skin first with alcohol and then with ether, to cover it with a thin layer of sterile vaseline. The wing vein was punctured with a fine scalpel and 15 to 20 c.c. of blood withdrawn in a Pasteur pipette. The blood was incubated for two hours and afterwards allowed to stand at room temperature for ten to twelve hours. Separation of the serum from the clot was greatly promoted by the use of the method suggested by Gardner (1917), of lining the test tubes in which the blood was incubated with a thin layer of sterile agar before use. The serum was centrifuged and then injected into the marginal vein of the ear. This operation is quite simple. As a general rule the does did not show any signs of ill health as a result of the injections. Within a few minutes of being replaced in their hutches they would be feeding and behaving quite normally. Two does, however, died during the series of injections. One of them—No. III.B. —was the smallest rabbit we used, and she died after receiving a dose of 7 c.c. —the largest dose given by us.
Though the does did not seem outwardly to suffer in health as a result of the treatment, it is probable that the treatment disturbed the normal metabolism. This conclusion seems to follow from the fact that out of 29 treatments given, 12 were not followed by the birth of any offspring. There are three possible explanations of such cases. The young may have been resorbed in utero, or they may have been aborted and consumed by the mother before the fact was observed. So far as either of these explanations is true, the cause is to be sought in the treatment to which the does has been subjected. Both Guyer and Smith, and Finlay (1923) found similar reduction in the fertility of treated animals. There is, however, a third possible explanation. Insemination may not have been always followed by fertilisation. Marshall (1922, p. 38) quotes Heape to the effect that if œstrus is experienced in winter by domestic rabbits, it may occur independently of the possibility of pregnancy. This latter explanation probably accounts for at least some of the failures.
The series of injections was usually begun between the eighth and the tenth day of pregnancy, and continued to about the twentieth day of pregnancy. The total amount given did not exceed 6 c.c. in some of the earlier cases ; later the dose was gradually raised until in one case it reached the total amount of 29 c.c. The interval between the doses was on the average three days. Full details of all the treatments given are shown in Table III. It will be seen that as a general rule each doe was treated with serum from the same fowl, though from time to time this rule was broken, and blood from different fowls was given to the same doe. One reason for this departure from the general scheme was that as a result of the continued withdrawal of blood the fowls were sometimes found to be losing weight. The fowls from which blood was being taken were regularly weighed and, when marked loss of weight was exhibited, withdrawal of blood was stopped until weight was regained. The does undergoing treatment were also regularly weighed, and the cases in which there were longer intervals than usual between the doses were generally due to the fact that injection was postponed because the doe was observed to be losing weight.
A reference back to Tables I. and II. enables the full meaning of Table III. to be appreciated. Table I. shows what treatment each fowl from which serum was taken had received, and Table II. shows the results of testing the serum of most of the fowls used. Thus the use of Table I. shows that of the 29 treatments given by us, 14 were with serum from fowls treated with rabbit lens, 5 with serum from fowls treated with ox lens, and 10 with serum from fowls treated both with rabbit and ox lens.
b. Results
No changes were ever observed in the eyes of treated does. A total of 57 young were obtained from the treated does ; details of all the litters are to be found in Table III. The eyes of all these young were very carefully examined in order to detect any abnormality that might exist. All the young rabbits which lived long enough to enable them to be handled were first examined without the aid of any instrument, and afterwards with an ophthalmoscope. The young rabbits proved to make admirable subjects for the use of this instrument. The examination of their eyes was very easy, after instruction in the use of the instrument had been kindly given us by Dr Adams. Guyer lays stress on the advantages of using albino rabbits owing to the ease of detecting lens abnormalities. This is doubtless so for gross examination, but any defects are detected by the ophthalmoscope as easily in pigmented as in albino animals. Any cloudiness, opacity, or other abnormality of the lens could not fail to have been detected. Some of the young did not live long enough to enable us to use the ophthalmoscope. A few were born dead or dying, and others died before they could be handled. The weakness of the young rabbits is presumably to be ascribed to the effect of the treatment upon the metabolism of the mothers to which reference has been made above. In such cases the young rabbits were examined immediately after death. Careful observations were first made to see whether any outward abnormalities were visible in the shape of malformations affecting the size, shape, structure, or prominence of the eyes. The eyes were then removed and the lenses dissected out. The lenses were examined for any sign of cloudiness or opacity. The two lenses were compared with one another, and frequently weighed to ascertain whether there was any difference in weight. The results of the examination by these methods of the eyes of all the young born from treated does may be summed up by saying that, with the exception of the two cases to be mentioned below, no sign whatever of any abnormality affecting the size, shape, or structure of the lens or of any other part of the eye was ever observed by us.
The first of these two cases refers to one of a pair of young obtained from doe No. X.B. As soon as this rabbit was observed, it was obvious that it was suffering from marked abnormality in the head region. It was clear from the outset, however, that whether or not the eyes were abnormal, the skull was considerably deformed. The upper jaw was twisted from right to left causing the upper incisor teeth to project at an angle. The left side of the face was sunk in and a flap of skin covered the left eye. When this rabbit was old enough it was carefully examined ; it died later when about three weeks to a month old. The skull was then macerated. The result of the examination of the living rabbit, of the dissection of the rabbit after death, and of the examination of the skull lead to the following conclusions. The abnormality centred entirely in the skull ; there were no other abnormalities apart from the displacement of certain parts due to the deformation of the skull. The sunken appearance of the left eye and its covering by a flap of skin were due wholly to the twisting of the bones of the skull. The left eye when dissected out proved to be in all respects normal. There were no abnormalities of structure, and the weight of the eye as a whole, and of the lens when dissected out, proved to be no less than those of the right eye. There would therefore seem to be no reason for attributing this particular abnormality to the treatment given to the mother.
The other case refers to one of four young obtained from doe No. IV.B. The lens of the right eye of one of these four young, when dissected out some hours after death, while quite normal in shape and size, was found to exhibit a pyriform opacity—the broader end of the opaque area pointing towards the centre of the lens. This observation must be interpreted in the light of other observations made by us. A series of observations has shown that rabbit lenses, both from treated and untreated parents, which, when perfectly fresh, are quite clear, often exhibit opacities a few hours after death. These opacities are sometimes transient—fading away within a few hours—while at other times they may remain visible for days. They usually become visible only after a certain period and increase with time. They are evidently the result of some post-mortem change. These opacities, however, so far as we observed them, were invariably spherical or ring-shaped, while the opacity in the case of the lens from the young rabbit referred to was pyriform. It was also rather more dense than what we may call the normal opacity. Our knowledge, however, of the formation of these opacities under normal conditions is so incomplete that the possible significance of this single observation cannot be accurately estimated. It would clearly be straining the facts to assume at the present stage that the opacity in this particular case was certainly due to the treatment to which the mother had been subjected, although it possibly may have been.
The results of treating the does with sensitised serum are therefore negative. One further experiment remains to be made with the material in our possession. The young from treated parents can be mated inter se and their offspring examined, and this is now in progress. Guyer and Smith obtained defective young from the mating of normal young of treated parents; and this further attempt is now being made to see if our material yields the same results.
3. Injection of Lens into Rabbits
a. Methods
The results obtained by Guyer, who records in another paper the injection of buck rabbits with rabbit sperm (1922a), suggested the experiments which will be described below. Guyer found that the sperm of bucks into which living rabbit sperm had been injected, was to some extent inactivated. It would thus appear that rabbits can form antibodies against their own sperm when it is injected directly into the blood stream. If antibodies against sperm can be so formed, it seemed likely that antibodies against lens could be formed in a similar manner. Guyer also injected rabbit lens directly into rabbits. After many failures and repeated trials (no figures are given) one treated doe gave one young rabbit with unaffected eye (Guyer, 1921). We also carried out some experiments along these lines. Rabbit lens or ox lens, removed and pulped under sterile conditions, was injected directly into the marginal vein of the ear. Three does were treated with rabbit lens and nine with ox lens. Ox lens was used in these experiments for the same reasons as it was used in the former experiments—that is to say, because it may be expected to have the same results as rabbit lens and because it is cheaper to procure. The operation was quite simple and similar to that of injecting fowl serum into rabbits. The details of all injections given are set out in Table IV. If the injected rabbits did not suffer from shock within a minute or so of receiving an injection, they did not appear to suffer in health in any way. Shock, however, frequently followed the injections, and out of twelve does injected six died in this way. The symptoms were similar to those exhibited by the fowls which suffered from shock on being injected with lens, and it is to be supposed that the shock was of the same nature in both cases. What its precise nature may have been remains undetermined. After four to six injections had been given, the doe was put into kindle as soon as the opportunity presented itself. One or two more injections were then given, which completed the series. This procedure was followed because it was assumed that, if antibodies were being produced at all by the treatment, the production of antibodies would be at its height at about the time when the lens was most vascular in the young in utero—that is to say, between the tenth and the fifteenth day of pregnancy.
b. Results
Experiments with precipitin tests were made to ascertain whether antibodies were, in fact, being formed in the blood of the does treated as described. Blood was drawn from the marginal vein of the ear. The ear was swabbed with alcohol and afterwards with ether, and then smeared with sterile vaseline. The blood was allowed to drip into a sterile test-tube. The subsequent procedure was exactly similar to that followed in testing the blood of the fowls injected with lens. The serum was separated, centrifuged, and tested in tubes containing dilutions of 0.4, 0.2, 0.08, and 0.04 respectively. The lens suspension was made up as before of 1 part lens to 39 parts normal saline.
The results of these tests are set out in Table V. Three does (Nos. V.B. VII.B. and VIII.B.) were tested against rabbit lens before any injections had been given. The results were negative. Later, does V.B. and VIII.B. were tested again after a series of injections of rabbit lens, and doe VII.B. after a series of injections with ox lens. V.B. and VII.B. both exhibited a reaction of the strength of “ + ” or “definite trace” in the tube containing the highest concentration of serum, while VIII.B. exhibited a fainter reaction in the same strength of serum. VIII.B. was again tested after further injections had been given—this time both against rabbit and ox lens. The same reaction was exhibited as before when tested against rabbit lens and a rather stronger reaction when tested against ox lens. A test with the serum from XIV.B. gave somewhat similar results. Tests with the serum of doe VI.B. gave peculiar results. After a series of injections with rabbit lens, a test against ox lens showed a reaction of the grade “ + ” or “definite trace” in dilution of the strength 0.2. A test made at the same time against rabbit lens became infected. After further injections another test was made both against rabbit and against ox lens. No precipitation was visible ; a distinct clearing effect was on the other hand clearly visible—the effect being most marked in the highest concentration of serum and decreasing progressively in the lower concentrations.
These results seem to point to the conclusion that an antibody is formed in the blood of rabbits against rabbit lens as the result of the injection of lens directly into the blood stream. It must be pointed out, however, that the evidence here adduced for this conclusion is slender. The experiments are few in number. Further, quite unlike the results of testing the serum of fowls treated with lens, when reactions of the grade “ + ” were visible in relatively low concentrations of serum, reactions of this grade were, except in one case, never visible except in the strongest concentration. More experiments are required in order to prove that antibodies against lens can be produced in this manner. But see Guyer (1921, p. III) and his references to others’ work (ibid., p. 113).
No effect of the treatment was ever observed in the eyes of the does which had been injected. Three litters totaling seventeen young were obtained. One litter of five was produced by a doe treated with rabbit lens, and two litters of four and eight respectively by does treated with ox lens. The young were examined precisely as described above, when giving an account of the manner in which abnormalities in the eyes of the young from parents injected with sensitised serum were looked for. No abnormalities of any kind were ever detected.
4. Discussion
We may begin by a brief comparison of our methods with those of Guyer and Smith.
The fact that we usually used intravenous, while they usually employed intraperitoneal injections into the fowls, is in favour of our having obtained a higher titre of antibodies.
The fact that all fowl serum used was previously tested for its precipitating power is in our favour ; since, although all our treated fowls developed some antibodies, we were able to choose those which had the highest titre.
While the amount of lens injected into fowls does not lend itself to direct comparison owing to the fact that we injected many of our fowls with a second or even a third series of doses in order to keep up the titre, it does not appear that the average amount of lens given by us was less than that given by Guyer and Smith.
If the first rabbits treated by us with fowl serum are excepted, the average dose given by us was about the same as that given by Guyer and Smith.
The number of treatments of does with sensitised fowl serum was in our series 29, in Guyer’s, 25. These produced 57 young as against 44 in Guyer’s work.
The number of treatments of does with direct injections of lens was in our series 12, as against “a large number” in Guyer’s. These produced 17 young, as against “a large number” in Guyer’s animals.
Two (= 6.9 per cent.) of our serum-treated does died; 3 (= I 2 per cent.) of Guyer’s serum-treated does died. In 12 of our treatments, no young were produced; in 9 of Guyer’s treatments no young were produced.
The use of agouti and Dutch pattern rabbits by us as against albinos by Guyer cannot well be supposed to have any influence on the success or failure of the treatment. It might have an influence upon the detection of positive results, unless the lenses were dissected out or examined with an ophthalmoscope according to our invariable procedure.
The use of ox lens instead of rabbit lens in many of our experiments might be supposed to make them less conclusive. However, since in every case but one, the reaction of the serum of a fowl treated with rabbit lens precipitated ox lens, or vice versâ, no objections can legitimately be lodged on this score.
We may now pass to a consideration of the theoretical problems at issue. The first point to stress is that the problem of inducing germinal modification by immunological methods remains in an extremely unsatisfactory state. Guyer and Smith with lens antibodies from the fowl have obtained 4 affected young out of 61 born alive, 3 of these from one mother.* With lens antibodies from the rabbit he obtained 1 affected young “after repeated trials” ; the number of young from the treated mothers is not given. The success in the fowl series was 6 per cent. of the young. In the rabbit series it was obviously much lower. Finlay (1923), with his mice, has obtained no successes (79 young, 47 parents); neither have we with our repetition of Guyer’s work (57 young, 29 parents). The percentage of successes has thus dropped still further. We agree with Guyer that it is not probable that his successes are due to accidental coincidence : but in view of the small number of cases we cannot regard it as wholly excluded, since the figures obviously do not reach the degree of certainty demanded, e.g., by the statistician.
What, however, we wish to affirm with some emphasis is this: that until (a) coincidence is excluded by a statistically significant number of successes ; and (b) something more is discovered as to the modus operandi of the antibody and in particular of the conditions of the failure or success of its action, it is not only unprofitable but illegitimate to attempt to generalise from the data and to apply broadcast the principles involved to evolutionary problems.
The danger of doing this is due to the fact that the general articles are seen and assimilated, but the. original data not critically examined; with the result that it is assumed that the general principles rest on a large body of well-ascertained fact. In support of this assertion we will quote from the recent utterances of no less an authority than Sir Arthur Keith (1923, p. 265) :—
“ … the well-known experiments of Guyer and Smith provide a rational explanation [of certain cases of coincidence between injury to the parent and defect in the same organ in the offspring]. They injected into the veins of doe rabbits, about the end of the second week of pregnancy, doses of a substance which has a selective and toxic action on the lens of the eye. Many of the young (italics ours) were born with defects of the eyes—cataract of the lens being particularly frequent. When these young rabbits grew up and bred, many of their young showed the same defects. The developmental disorder could be transmitted in the spermatozoa as well as in the ova. These experiments show that the germ-plasm can be reached from without….” (italics ours).
Guyer himself has not hesitated to build large super-structures of theory on his facts. It is true that in the closing paragraphs of his papers he has been careful to add the most cautious qualifications ; but these have not appeared to check his speculations’ flight.
We will confine ourselves to one quotation, e.g. 1922c, p. 126 :—
“ It is reasonable to suppose that if an animal’s own tissues became displaced, injured, or otherwise modified, they might cause the production of antibodies…. With the occurrence of injury, under-stimulation or pronounced changes in any part of the body, serological changes would probably be produced in the blood-stream…. Such a hypothesis affords, perhaps, a plausible explanation of such deteriorative evolutionary processes as those seen in the formation of vestigial organs.”
He further (loc. cit., pp. 128-9) attempts a serological explanation for constructive evolutionary (germinal) change under the influence of recurrent somatic hyperplasia, although in a much more tentative fashion.
There are some further points which demand discussion. The explanation given by Guyer and Smith obviously demands that there shall be either identity or considerable resemblance between the proteins of the lens and the proteins of one at least of the genes concerned in lens-formation. They rightly stress the point that the two materials need not be identical ; but it is clear that the lens-protein must be more like the lens-gene-protein than the protein of any other gene, for otherwise the effect would not be confined to effects dependent upon lens-damage.
There is, of course, nothing impossible in the idea that certain tissues of the adult are produced essentially by the localised multiplication of certain molecules of the germ. But we must remember that there are many known cases where this cannot well be so. What chemical correspondence can there be between white eyes in Drosophila and the gene concerned? Or, still more, between the characters reduplicated thorax or forked bristles and their genes? In such cases, then, the genes appear to modify the course of chemical reactions which are themselves the outcome of a long chain of other reactions already gone before in ontogeny : not only need there not be any chemical correspondence between the first and last links of the chain, but it is hard to imagine any.
It may be objected that many genes might thus exert a modifying action ; but that what they modified was a chemical skeleton, so to speak, which depended upon the multiplication of genes of the same chemical character as the organs they finally determined. There is again nothing a priori to be said against such an idea. Recent discoveries, however, in the field of developmental physiology make it difficult to understand how such a mechanism should operate. Spemann (1921) has recently discovered that the formation of the axial organs (medullary plate in particular) in Amphibia depends upon the presence of a piece of the dorsal lip of the blastopore, which acts as a “differentiator.” If an extra “differentiator,” even from another species, be grafted into a late blastula, it will induce the formation of an extra medullary plate from the host’s tissues : the induced medullary plate later proceeds on a normal course of differentiation.
What is more, the differentiator is indirectly responsible for the formation of the organ which here concerns us, viz., the lens; for the lens appears in a definite position with reference to the embryonic eye, which again was determined almost at once within the early medullary plate. The position of the lens is always determined in this way, whether it is self-differentiating or dependent upon the optic cup.*
In the present state of our knowlege it appears simplest to suppose that the differentiator, which is the region of highest activity in the germ, exerts its action in virtue of the stimulating effect of this activity itself upon the surrounding tissues ; while the different chemical reactions occurring in the different regions of the future medullary plate region under its action depend upon their differing positions in the axial gradient systems of the germ. Detailed consideration is impossible here, but reflection will show that it is very difficult to imagine the transmitted stimulus from the differentiator bringing about in one region the multiplication of genes containing proteins like those of nervous tissues (for brevity’s sake we may say “genes with nerve-tissue protein”); in a second, that of genes with retinal-tissue proteins; in a third of tapetum-tissue proteins; in a fourth of lens-proteins. It looks much more as if a chain of reactions was in each case set going which moved toward an appointed end, but an end not necessarily resembling any of the substances present at the beginning.
In brief, Guyer’s hypothesis implies a considerable degree of preformation—a position away from which both experimental embryology and genetics have been steadily moving as discovery corrected their first and cruder notions.
Finally, it remains to mention the one apparently incontrovertible case of adaptive modification of the germ-plasm, in order to show how rare and difficult, even with the most favourable material and methods, such modification is.
Jollos (1921), in a monumental series of researches, has attempted to induce permanent adaptive modifications of Paramecium in respect of resistance to heat and to poisons. Both treatments gave essentially similar results. Thousands of individuals of pedigree stock were exposed to high temperature, and to solutions of arsenious acid and calcium salts in various ways. Each stock started with a sharply-marked upper limit of resistance, differing for different stocks. The only treatment that produced a marked heightening of resistance was the placing of the animals in temperatures or solutions above the permanently lethal limit, but for sub-lethal periods. In this way, strains were produced whose resistance was very materially heightened (e.g. from resistance to 1 per cent. arsenious acid to nearly 5 per cent.).
However, the resistance thus induced was never permanent. In some cases it disappeared at the first conjugation period, in others before it. In still other cases, although it persisted through one and even two conjugations, it eventually dis-appeared spontaneously. Endomixis often reduced the degree of resistance. It should be noticed that the changes were usually but not always adaptive. High resistance could only be induced by repeated treatments ; each successful treatment increased the resistance by a definite and constant amount. When spontaneous disappearance of resistance took place in such stocks, the important fact was noticed that it also occurred by steps, and by the same steps by which it had increased, only in reverse order, as if these represented positions of equilibrium. Jollos not unreasonably concludes that modifications of resistance which disappeared with conjugation were due to changes in the macronucleus, whereas those which were able to persist through conjugation were due to changes in the cytoplasm.
The only changes in resistance which showed themselves to be permanent were changes induced by treatment during conjugation. These he ascribes to changes in the micronucleus, i.e. in the germinal constitution. He only, however, obtained a few such cases out of many trials.
This work is of the greatest importance. It shows us (1) that adaptive changes lasting over many generations may be induced ; (2) that these changes, however, need not be hereditary sensu stricto ; (3) that permanent, truly hereditary changes, also adaptive in character, may be induced ; (4) that, however, they are very difficult to induce, and occur with an extremely low frequency; (5) that there is no reason to suppose that they are the result of longer persistence of the agencies which produce the non-hereditary but long persistent modifications, but rather the result of applying the agencies when the germ-plasm is in a special state. From this last point of view, the recent work of Sturtevant and Morgan (1923), who show that in Drosophila mutation is, in the case of the reverse mutation bar to normal, associated with crossing-over at or near the locus of the gene which mutates, is of considerable interest.
The important work on the harmful heritable effects of alcohol (Stockard and various others ; see e.g. Stockard, 1916, 1922), and of X-rays (see Little and Bagg, 1923) need not concern us here, since agencies of a general noxious character are employed, and the effects produced are again of a general nature. In contradistinction to this, the effects claimed by Guyer and Smith, by Jollos, and by Sturtevant and Morgan, are all specific effects (or mainly so—some of Jollos’ being of a more general character), in spite of the fact that both their nature and the methods involved in their production are so different in the three cases.
Such work as that of Jollos, and of Sturtevant and Morgan, puts the problem on a different level. For the often academic discussion of the “Inheritance of acquired characters,” we are given the problem of the experimental modification of a germ-plasm about whose characteristics, thanks to the researches of Johannsen and of the Mendelians, especially of the Morgan school, we know a good deal. I.e., the problem has had its limits defined, and is removed from one dealing primarily with adult characters to one dealing primarily with genetic factors and with the physiology of their development.
We do not pretend that the facts presented in this paper are in any sense decisive. We are only concerned to point out that we embarked upon the task of repeating Guyer’s work, and that, although the technique is not difficult, we have been unable to do so. We therefore hope that other investigators, especially if they have facilities for work on a larger scale than was possible to us, will attempt the solution of the problem. It may be worth while suggesting that an animal with a longer period of pregnancy might offer more favourable prospects.
For the present, we must claim that the serological induction of hereditary modifications is not proven, although Guyer has made it probable that it may occur in rare cases ; and especially that any far-reaching speculations based on the assumption of this induction are at present unjustified.
Acknowledgements
Our acknowledgments are due to the Royal Society, whose allotment of a Government grant to us, alone made the work possible. Further, to Professor G. Dreyer, F.R.S., and Dr A. D. Gardner, F.R.C.S., of the Department of Pathology, without whose constant advice and ready help we should often have been in difficulties. Part of the work was carried out in the Department of Pathology ; the major part in the Department of Comparative Anatomy, to whose head, Professor Goodrich, we also wish to express our thanks for the facilities and encouragement given to us.
References
Guyer claims 4 markedly affected, 5 “somewhat affected.” In view of the slightness of the symptoms, and of his description, and of the absence of further discussion of these 5 animals with “watery eyes,” we feel justified in excluding them from the certainly positive cases.
The explanation of the fact that even in two closely-related species, the lens in one shows dependent differentiation, in the other self-differentiation, is very possibly to be sought in a simple difference of time-relations, the so-called self-differentiation being only a more precocious dependent differentiation.