Studies on the shape of the Golgi Apparatus. II. Observations on the fresh Egg of the Indian Earthworm, Pheretima posthuma

In the first paper (1929) of this series on the egg-follicle of Culex fatigans I have shown that the Golgi elements in the oocytes, in the nurse-cells, and in the follicular epithelial cells are in the form of vesicles, each vesicle showing an osmiophilic or argentophilic rim and an osmiophobic or argentophobic central area. The contents of the Golgi vesicles of the nursecells and of the follicular epithelial cells remain non-fatty throughout the growth period of the egg-follicle; while those of the majority of the oocyte vesicles become fatty, although they do not swell up. If a fresh egg is ruptured and its contents studied under the microscope the Golgi vesicles are seen performing a very interesting dancing movement while the proteid yolk bodies are stationary. The presence of fat inside the interior of the Golgi vesicles raises their refractive index and enables one to study them in fresh eggs with the utmost ease. Vesicular Golgi elements have also been demonstrated by the writer (1928) in the eggs of a spider, by Nath and Husain (1928) in the eggs of a Scolopendrid, by Nath and Mehta (1929) in the eggs of Luciola, and by Nath and Piare Mohan (1929) in the eggs of the cockroach. It is important to note that in all the above cases the Golgi elements were traced from the youngest to the most highly advanced oocytes, and in some cases (Culex and Luciola) vesicular Golgi elements were observed even in the undifferentiated primordial germ-cells. In the case of the cockroach such elements were found even in the terminal thread of the ovary. The above conclusions were based not only on the study of fixed preparations, which by themselves are entirely worthless at least for the study of the morphology of the Golgi apparatus, but also on the study of fresh cover-slip preparations treated with vital dyes or with 2 per cent, osmic acid for a very short time.

The Golgi element in the eggs then is a vesicle consisting of two fundamentally different substances, namely, the osmiophilic peripheral rim and the central clear osmiophobic substance. The writer has always been puzzled by the ‘crescents’ or the ‘dictyosomes’, forms under which the Golgi elements have been constantly described in the male germ-cells by such competent writers as Gatenby and Bowen. But a complete and final solution of this puzzle has unexpectedly come in the form of an excellent publication on the Golgi apparatus and vacuolar system of Cavia, Helix, and Abraxas by Gatenby (1929). This writer has shown that even in the male germ-cells there is a number of neutral red-staining vacuoles lying near, but not attached to, the Golgi dictyosomes. In some cases (e.g. Helix) the vacuoles are in the archoplasmic area at the periphery of which the Golgi dictyosomes lie, while in others (e.g. Cavia, Saccocirrus, and Abraxas) they are extra-archoplasmic. These vacuoles are distinct from the Golgi dictyosomes not only in the ‘resting’ cells but also during the meiotic division when they are sorted out roughly into two equal parts. It is very important to note that, according to Gatenby, these vacuoles are not consistently argentophilic, whereas the dictyosomes are. Gatenby also gives a useful review of the recent work of Hirschler, Monné, and Voinov on the male germ-cells of lizard and grasshopper, Cerithium and Helix, and Notonecta respectively. These authors, as quoted by Gatenby, have also shown a vacuolar system lying near, but distinct from, the Golgi dictyosomes.

Parat’s views on the homologies of the cell-constituents of the Helix spermatid, which are well known, are diametrically opposite to those of Gatenby, Hirschler, & c. Parat considers the archoplasmic vacuoles as the Golgi apparatus and the cortical dictyosomes as the modified mitochondria or the ‘lepidosomes The evidence brought forward by Gatenby and others, however, seems to be overwhelming, and one cannot have any hesitation in accepting their views.

The paper of Gatenby cited above lends strong support to my views on the shape and function of the Golgi elements in oogenesis. I have consistently described the Golgi elements in eggs as vacuoles or vesicles, each vesicle having an osmiophilic or argentophilic rim. According to Gatenby, there exists in the animal cell a vacuole or a system of vacuoles primitively associated with, and probably produced by, the argentophil cortex of the Golgi apparatus. Further, in the eggs the vacuole is ‘closely related to the chromophil substance of the Golgi apparatus’. ‘The work of Miss A. G. Hill in this laboratory has convinced the writer that in such examples of oogenesis as that of Daphnia, the Golgi element is a cortex on the vacuole, and the division of the element brings about a division of the associated vacuole.’ In showing that the mysterious ‘idiosome’ or the ‘sphere substance’ which in many cases appears as the chromophobe part of the Golgi element is nothing but a collapsed vacuole, Gatenby has made a first-class contribution to cytology.

It appears, therefore, that the essential part of a Golgi element is the peculiarly osmiophilic or argentophilic material which in the somatic cells occurs in the form of granules, rods, or a reticulum, while in the plant-cells it is found, as claimed by Bowen (1928) and by Patten and collaborators (1928), in the form of platelets which impregnate with osmic acid like the typical Golgi apparatus. Bowen has insisted that the plant vacuole is not osmiophilic, that is to say, it is a mere vacuole without the osmiophilic rim; and since Gatenby, Hirschler, & c., have shown a vacuolar system distinct from the true Golgi apparatus in the male genn-cells, one is driven to the conclusion that the theory of Guilliermond and Parat homologizing the plant vacuole with the Golgi apparatus cannot be any longer upheld.

If the plant vacuoles and the vacuoles of the male germ-cells are mere vacuoles and not the Golgi apparatus, it perhaps becomes doubtful if the neutral red-staining vacuoles which Parat and his collaborators have been describing in eggs are surrounded by a truly osmiophilic Golgi material. At any rate Nath and collaborators find that in the eggs of the spider (Nath, 1928), Culex (Nath, 1929), Pheretima, medicinal leech (Nath, unpublished), Scolopendrid (Nath and Husain, 1928), Luciola (Nath and Mehta, 1929), and the cockroach (Nath and Piare Mohan, 1929), the Golgi vesicles are not stainable, at least brilliantly as claimed by Parat, with neutral red even in the early stages of oogenesis when in some of the above cases (spider, Scolopendrid, Culex, and the cockroach) their contents are non-fatty.

It is essential to discuss the remarkable papers of Foot (1894, 1896, and 1898), and of Foot and Strobell (1898, 1900, and 1901), on the cocoons and eggs of Allolobophora faetida. I consider this work of a very high order because it is based on an extensive study of fresh cover-slip preparations, and the authors have taken pains to publish a large number of excellent photographs. In those days very little was known about the mitochondria, and the Golgi apparatus was entirely unknown by zoologists. In spite of this the authors have given a very faithful account of oogenesis in this earthworm.

These authors (1901) described two types of granules in the egg of Allolobophora, namely, ‘deutoplasmic’ or ‘osmiophile’ granules and smaller ‘archoplasmic’ granules which arise from the disintegration of the ‘yolk-nucleus’ (Calkins, 1895). The osmiophile granules are present long before the disintegration of the yolk-nucleus, and they can be demonstrated during every stage of development of the egg. They are found in nearly all the cells of the ovary, from the small cells near the proximal end, which show the first indication of a yolknucleus, to the large oocytes at the distal end (Foot and Strobell, 1901, p. 518). These granules can be demonstrated during all stages of the growth of the egg, the maturation, fertilization, and cleavage (p. 519). The yolk-nucleus granules fail to blacken with osmic acid, even after many hours’ immersion in a 1 percent, solution, whereas an immersion of fifteen or even five minutes is sufficient to blacken the osmiophile granules. The yolk-nucleus invariably stains intensely with haematoxylin, while the osmiophile granules very rarely react to this stain (p. 524). The osmiophile granules also show conspicuously in the living egg, and agree both in size and position with those seen in fixed material. They appear in the substance between the cells and in the small cells (oogonia ?) near the proximal end of the ovary (p. 531). In the oogonia and in the oocytes of all stages the yolk-nucleus is morphologically an accumulation of granules. When the granules are scattered through the cytoplasm they are very difficult to demonstrate, but when they are aggregated into more or less definite masses they can be readily differentiated (p. 522).

It is convenient to anticipate at this stage that the ‘osmiophile’ granules and the ‘archoplasmic’ or ‘yolk-nucleus’ granules are the Golgi elements and the mitochondria respectively.

Harvey (1925), working on the egg of Lumbricus terrestris, however, arrived at the following conclusions:

1. The yolk-nucleus is merely a mass of mitochondria.

2. The mitochondria arise as a cap of threads over the nucleus.

3. The mitochondria are not clearly defined in the very young oogonia.

4. The Golgi apparatus consists of numbers of Golgi elements lying separate in the cytoplasm. There is never any attempt at concentration of these elements round one central mass.

5. The Golgi elements are probably little platelets or spheroids somewhat resembling blood-corpuscles in shape. They are not rods. As fixed by Da Fano technique, each element is a little plate with a very lightly impregnating centre and a very heavily impregnating rim.

6. The Golgi elements may probably arise from the cytoplasm.

7. Yolk is present, and probably arises from the cytoplasm.

8. No direct metamorphosis of either mitochondria, Golgi apparatus, or nucleolus into yolk was observed.

Unfortunately these conclusions were based entirely on the study of fixed preparations because ‘the fact that the egg of Lumbricus is full of highly refractive granules and globules of yolk, fat, &c., which obscure any signs of the fine threadlike mitochondria, makes the study of the inclusions in the living cell impossible

Later, Gatenby and the writer (1926) published a paper on the same subject, but they disagreed with all the statements of Harvey except those contained in paragraphs 1 and 8. Their results were based not only on the study of fixed preparations, but, which is very important, on the study of fresh cover-slip preparations made by Gatenby alone, as the writer was compelled to leave Dublin for his home in June 1925. On his return to India he took the earliest opportunity of making much more extensive intra vitam and other observations on the egg of the common Indian earthworm. Indeed he has been demonstrating the mitochondria and the Golgi elements to his Honours School classes with the utmost ease, not only in the eggs of the earthworm but also in those of the medicinal leech. The cytoplasm of both these eggs is not encumbered with yolk, the only inclusions present being the Golgi elements and the mitochondria. For this and other reasons to be mentioned subsequently these two types of inclusions stand out prominently, making these objects truly classical for observations on living material.

The conclusions arrived at by the writer with regard to the oogenesis of Pheretima were substantially similar to those arrived at by him and Gatenby in the case of Lumbricus. In spite of the fact that he discovered important additional facts in support of the previous conclusions on Lumbricus (1926) and also in support of his views, which he holds very strongly, on the shape and function of the Golgi elements in eggs in general, he preferred not to go to press, hoping that the differences would automatically disappear when it would be possible for Harvey to observe living eggs. That hope, however, has been only partially fulfilled as the following quotation will show (Harvey, 1929): ‘The author would like to acknowledge here the correctness of Professor J. Brontë Gatenby’s statement that the egg of Lumbricus can be easily studied in the living condition. He can only attribute his former failure with it to lack of experience in this particular field. Further observation has demonstrated to him that the Golgi bodies can be easily seen in these eggs, living and unstained. With Gatenby’s further statements about this egg, however, he still remains in disagreement, and is strengthened in his own opinions by the re-examination of numerous samples. The author holds that the mitochondria are thread-like, not granular, that there is a certain amount of material, yolk for want of a better word, in the eggs, and he is not convinced that only one Golgi body is present in the oogonia and youngest oocytes.’

The above lines appear as a footnote in a paper on the oogenesis of the crab. The writer would have greatly welcomed a statement from Harvey with regard to the shape of the Golgi elements in the eggs of the earthworm which he now claims to have studied in the living condition, because the morphology of the Golgi element is of fundamental importance, particularly in eggs, and is inextricably bound up with their function.

It has already been mentioned that the earthworm ovary may be considered a very favourable object for the study of the mitochondria and the Golgi elements in the living condition without the help of any vital dyes. This is true, strange as it would appear, especially in the case of the Golgi elements which stand out much more prominently than the mitochondria in the freshly extirpated ovary. If the ovary is mounted flat on the slide in a drop of normal saline so that the oocytes are not allowed to cover each other, one can see the oocytes arranged in filaments radiating from the septal insertion of the ovary. The oogonia are proximal, that is to say, they lie near the septal insertion and the oocytes lie more distally, the most highly developed oocyte lying at the free end of the filament. On account of this peculiarly favourable arrangement the study of different stages in one specimen only is greatly facilitated.

In text-fig. 1 is shown a group of four oogonia in which there is not a trace of any granules which can be assigned to the category of mitochondria, the cytoplasm having a hyaline glasslike appearance. The Golgi elements, however, stand out very prominently as highly refractile spherules of a dark-greyish colour. In one oogonium in Text-fig. 1 there is a single Golgi element, while in others there are four or five such elements. I have frequently seen oogonia with one, two (Text-fig. 9 a), four, or more Golgi spherules; but I have never observed three such spherules in - an oogonium, strange as it may appear, although my studies have been spread over many years and numerous specimens. With the growth of the oogonium the Golgi spherules spread out in the cytoplasm (Text-figs. 2, 8, 4, 5, and 6) till in an advanced oocyte (Text-fig. 6 a) they are distributed more or less uniformly throughout the egg. In the oocytes the Golgi elements perform a dancing movement which is very interesting to watch.

The mitochondria appear for the first time either as a thick mass (Text-figs. 4 and 9b) or as a horseshoe-shaped structure closely fitting the nuclearmembrane(Text-figs.3,9 c and d). Very soon the mitochondrial cap breaks away from the nuclear mem brane and is quickly resolved into aggregations of mitochondrial granules (Text-fig. 5). The mitochondrial aggregations now become loosened (Text-fig. 6) till in advanced oocytes the granules are distributed more or less uniformly throughout the cytoplasm (Text-fig. 6 a). The mitochondria, even in the earliest cap, are in the form of very small, whitish granules with a refractive index much lower than that of the Golgi elements. They perform a very rapid dancing movement closely simulating the similar movement of bacteria. The very much smaller size of the mitochondrial granules, their lower refractive index, their peculiar distribution in the egg, and their whitish appearance as compared with the dark-greyish colour of the Golgi spherules, are factors which preclude any possibility of confusion between these two kinds of cytoplasmic inclusions.

The nuclear contents consist of faint and fine fibres with a prominent semi-solid and refractile nucleolus. Near the septal insertion of the ovary some oogonial nuclei can be seen in the prophase with much thicker and more conspicuous fibres inside them.

When some time has elapsed the egg begins to disintegrate under the eye of the observer, and it breaks down into fragments which seem to be surrounded by a thin limiting membrane. In Text-fig. 7 such a fragment has been shown. Most of the Golgi elements are crowded in one corner and are perfectly stationary, but the mitochondrial granules which occupy the major portion of the fragment, with a few Golgi elements lying amongst them, perform a very rapid dancing movement which contrasts prominently with the stillness of the crowded Golgi elements in the corner.

In other cases at the time of the death of the egg the eggmembrane ruptures at one or several places and its contents flow out in the form of a rapid stream containing the mitochondrial granules and the Golgi elements (Text-fig. 8). They remain unaltered for a long time after the death of the egg, showing that they have a firm consistency and are not so susceptible to injury as is generally supposed. The Golgi vesicles of the Culex egg also remain unaltered for a long time after the death of the egg. Vital dyes, namely, neutral red and janus green B, have been extensively used in the form of very thin watery solutions, but they do not in any way improve the appearance of the egg, if indeed an improvement were desired. The Golgi spherules do not stain at all with either of these dyes, but the mitochondria do take up very slightly a blue colour with janus green B.

Elsewhere (1928) the writer has laid stress on the value of treating the eggs for a short time only with 2 per cent, osmic acid for determining the shape and the nature of the chemical contents of the Golgi elements. It is true that the cell is killed by a short exposure to osmic acid, but the appearance of such a cell is almost like that of the living. This is borne out by the clear-cut experiments of Strangeways and Canti (1927), who have shown that the cell after a short period of fixation with 2 per cent, osmic acid is almost like the living cell in vitro with regard to all the inclusions.

The first change to be noted in the appearance of the Golgi apparatus after a short period of five to ten minutes’ immersion in osmic acid is that they look copper-coloured, but they still appear solid as they do in the living egg. After half an hour’s treatment with osmic acid, however, a very important change comes over the appearance of the Golgi elements, a change that gives a definite clue to their morphology. Each element now shows a dark rim and a lighter central area (Text-fig. 9). From this one has to conclude that the Golgi element is not a solid or semi-solid body, but is really a vesicle with a definite osmiophilic rim and a hollow interior.

Experiments with the centrifuge were performed to prove the entire absence of albuminous ‘yolk-droplets’ which, according to Harvey, are present in the egg of Lumbricus. The centrifuge used in these experiments was an ordinary hand

[ufig]

centrifuge fitted up with a motor, giving about one thousand revolutions per minute. In Text-fig. 10 are shown five oocytes centrifuged for thirty-five minutes and kept in osmic acid for two hours. The mitochondria and the Golgi elements are completely separated into two layers, the former appearing as small yellowish granules and the latter as vesicles, each vesicle showing a definite osmiophilic membrane and a clear central substance represented by the white background of the paper. In Text-fig. 11 are shown three oocytes centrifuged for half an hour and kept in 2 per cent, osmic acid for ten minutes only. The mitochondria appear as slightly yellowish granules, but the Golgi elements look solid and copper-coloured as their rim is not yet impregnated. If there were any yolk-droplets of an albu-minous nature as described by Harvey it is inconceivable that they should be missed in a centrifuged egg treated with two per cent, osmic acid, for they would as usual stand out prominently as solid, white, or slightly yellowish bodies much bigger in size than the tiny mitochondrial granules and distinct from the blackened Golgi vesicles.

The difficulties in the way of getting first-rate preparations of the earthworm ovary are considerable. This material is remarkably refractory, and whatever the fixative used there is always a certain amount of shrinkage as shown by the tearing away of the cytoplasm from the egg membrane—a fact emphasized by Foot and Strobell (1901). Nevertheless I have been able to obtain satisfactory preparations of eggs, except the most highly advanced which lie at the distal ends of the ovarian filaments. It is almost impossible to fix these latter in a satisfactory manner. Foot and Strobell, who made very extensive studies of the ovarian eggs of Allolobophora as well as of those contained in the cocoons, ascribe this refractoriness to the fact that a large number of eggs at the distal end of the ovary degenerate and are unable to develop. Luckily, however, the egg of the earthworm is such an ideal material for intra vit am observations that fixed preparations can perhaps be entirely dispensed with except for purposes of control.

In the course of my experiments I have discovered a point of considerable importance (vide infra), namely, that the Golgi elements of the earthworm which are vesicular in nature contain a certain amount of fat inside them. This point has been overlooked both by Harvey (1925) and by Gatenby and Nath (1926). When the whole ovary is mounted after twenty-four hours’ fixation in Champy’s fluid the Golgi vesicles appear as very black solid granules not only in the oocytes but also in the youngest oogonia (Text-fig. 12). If the same preparation is studied after about a month one is surprised to find that the Golgi elements appear as colourless or at most slightly greyish bodies. This is undoubtedly due to the decolorizing action of the xylol in the Canada balsam. When the paraffin ribbon containing Champy-fixed sections is studied the Golgi elements appear as blackened granules; but immediately after exposure to xylol to remove the paraffin the elements are decolorized and appear as colourless or slightly greyish bodies (Text-fig. 13), which are liable to be missed unless the condenser of the microscope is considerably lowered to reduce the amount of light. In whole mounts, however, the xylol of the Canada balsam will naturally take a much longer time to decolorize the blackened Golgi elements. The above observations adequately explain the statement of Gatenby and Nath (1926) that ‘the Golgi elements, however, are slightly osmiophile but do not go black’.

‘Kolatschev’ preparations not only reveal the vesicular nature of the Golgi elements, but also adequately explain certain forms which these elements may sometimes assume. The optimum time for the proper impregnation of the Golgi elements of the earthworm egg is about four days’ incubation at the temperature of about 35° C. The mitochondrial granules are usually blackened in the regions of close aggregations, but they can be very successfully bleached by potassium permanganate followed by oxalic acid.

When an optimum impregnation and proper bleaching have been secured the Golgi elements appear vesicular, each vesicle showing an intensely blackened rim and a much lighter central area (Text-fig. 16a). But more often than not the Golgi elements assume different forms, all of which must be interpreted as artifacts. In Text-fig. 165 and c the Golgi elements are more or less ellipsoid and rhomboid respectively—appearances which are due to the shrinkage inevitable in prolonged osmication. When the incubation has been carried out for about six days most of the Golgi elements appear uniformly jet-black (Textfig. 16d, e, and f). Even after shorter periods of incubation (four days) most of the Golgi elements appear solid (Text-figs. 14 and 15). It must be clearly understood that the ‘solid’ appearance of the Golgi elements after fixation in ‘Kolatschev’ is due not only to the blackening of the fatty contents of the vesicle, but also to the heavily impregnated rim, whereas the same appearance of the Golgi element after twenty-four hours’ immersion in Champy’s fluid is due solely to the reduction of osmic acid by the fat, the rim of the vesicle being not at all impregnated. It is due to this fact that the Golgi vesicles which are ‘solidified’ after fixation in ‘Champy’ are so quickly decolorized by xylol or turpentine, whereas after ‘Kolatschev’ these reagents fail to decolorize the Golgi elements even in months.

After a light impregnation with ‘Kolatschev’ the Golgi elements assume still different appearances. In Text-fig. 16 i the rim of the vesicle has completely missed impregnation and the blackened fatty contents have been decolorized by xylol. Consequently the vesicle appears as a colourless or slightly greyish body, reminding one of the so-called ‘yolk-droplets’ of Harvey. In some cases, however, the rim of the vesicle is very lightly impregnated (Text-fig. 16 h). The form of the Golgi element, which I cannot definitely explain, is a heavily impregnated crescent with an osmiophobic substance attached to it (Text-fig. 16 g). This form results either from a partial blackening of the rim, or it represents an optical section of the vesicle.

The greatest amount of shrinkage of the egg takes place after ‘Da Fano’, but the Golgi elements are heavily impregnated and almost invariably appear as solid blackened granules. The mitochondria appear as much smaller granules, golden in untoned preparations and colourless or slightly greyish in the toned ones.

The oogenesis of the earthworm is the simplest at present known. The only inclusions present in all stages of oogenesis are the mitochondrial granules and the slightly fatty Golgi vesicles. There is no yolk in the egg either fatty or albuminous. This fact is at once understood when it is remembered that the egg is nourished by an albuminous material present in the cocoon.

Thirty years ago Foot and Strobell, working with a technique much less efficient than the technique of the present day, described these inclusions in a remarkably accurate manner. A reference to the section on the previous work on the earthworm ovary will at once make it clear that the ‘osmiophile’ granules and the ‘archoplasmic’ or the ‘yolk-nucleus’ granules of Foot and Strobell are the Golgi vesicles and the mitochondria respectively.

According to Harvey (1925 and 1929) the mitochondria arise as a cap of threads over the nucleus, and this cap grows in size and density, migrates away from the nuclear membrane and breaks up into its component mitochondrial threads, which become evenly spread throughout the cytoplasm of the cells. In by far the greater number of eggs of different animals studied with the most modern technique the mitochondria have been described and figured as granules. In very few cases of oogenesis have filamentous mitochondria been described. For instance, Hibbard (1928) describes filamentous mitochondria in the egg of Discoglossus. Indeed, so far as I am aware, this egg and that of Lumbricus as described by Harvey are the only eggs which are stated to have mitochondria in the form of filaments. It is also true that in tissue culture of somatic cells the brothers Lewis (1914 and 1915) have described rapid changes in the shape of the mitochondria. The granules may become arranged in a linear series or the individual granules be converted into filaments which can give rise to complicated networks.

In the egg of the earthworm, however, the mitochondria are definitely granular, both in the oogonia and the oocytes. This has been shown by Calkins (1895) and by Gatenby and Nath (1926) in the case of Lumbricus, and by Foot and Strobell (1901) in the case of Allolobophora. I am at a loss to understand the statement of Harvey (1929) that even in the fresh eggs the mitochondria are thread-like and not granular, and I am compelled to ascribe this statement to faulty technique or observation. Harvey’s mitochondrial ‘filaments’ in fixed preparations are certainly due to the artificial alignment of granules. A similar artifact is produced in the egg of the red cotton bug. In the fresh egg and after the mitochondrial technique, namely, Flemming without acetic and iron haematoxylin, and Champy-Kull and acid fuchsin, the mitochondria appear as granules, but in some ‘Kolatschev’ preparations they appear as ‘threads’.

According to Harvey ‘a fair amount of yolk is present in the developing egg of Lumbricus as quite large droplets. Although not universally present in the oogonial and young oocyte stages, yet these droplets are to be found in all the stages, and it is significant that they are present in the young oogonium before the mitochondrial cap can be detected with certainty.’

‘They do not go red in Champy-Kull or Bensley-Cowdry, they are yellowish or brownish in most chrome-osmium preparations, are easily decolorized by turpentine after Mann-Kopsch, do not disappear after Carnoy fixation, &c.’

Foot and Strobeli, and Gatenby and Nath did not find any yolk-droplets in Allolobophora and Lumbricus respectively. I do not find them in the case of Pheretima, the only inclusions being the mitochondrial granules and the Golgi vesicles. If albuminous yolk-droplets were present, they should have been thrown in a separate stratum in the centrifuged eggs. What Harvey describes as yolk in his Mann-Kopsch preparations are really some of the Golgi vesicles which have not been properly impregnated by osmic acid (see Observations). What he describes as yolk in his chrome-osmium, Champy-Kull, and Bensley-Cowdry preparations are again the Golgi vesicles which have not been properly stained. Indeed it is not easy to stain them after treatment with the above fixatives on account of their vesicular nature. It is difficult to stain the rim, and unless the stain is precipitated inside them, they fail to stain. Foot and Strobell also found that the osmiophile granules very rarely react to haematoxylin (vide supra). The fact that ‘yolk-droplets’ do not disappear after Carnoy fixation does not prove that they are true albuminous yolk. In my Bouin preparations of the Pheretima egg granules answering to the description of ‘yolk-droplets’ of Harvey appear, which are really the distorted Golgi elements. Every experienced cytologist knows that the mitochondria and the Golgi elements are not always completely washed out by fixatives containing acetic acid. In many cases they remain in a very much distorted condition. I am reminded of the ‘metaplasm’ or ‘formations ergasto-plasmiques’ of earlier workers on the spermatocytes of L i t h o b i u s, which are really the very much corroded Golgi elements as shown by Nath (1925). Lastly, it is certainly unusual that ‘yolk-droplets’ should be present in the oogonia.

I am also unable to support the statement of Harvey that ‘the Golgi apparatus consists of numbers of Golgi elements lying separate in the cytoplasm and that they ‘may probably arise from the cytoplasm I have often seen the Golgi vesicles lying very close to each other up to the stage when they are only four in number (Text-figs. 1 and 12). From this stage onward they start dispersing, till in advanced oocytes they are more or less uniformly distributed throughout the cytoplasm. It is impossible to be certain with regard to the origin of the Golgi elements, but the fact that in the very young oogonia (Textfigs. 1, 3, 4, and 12) they lie near each other points towards the conclusion that at least in the early stages new Golgi elements rise from the division of the pre-existing ones.

Harvey (1929) is ‘not convinced that only one Golgi body is present in the oogonia and youngest oocytes’. In reply, in addition to inviting his attention to Text-figs. 1 and 12 of this paper, in which only one Golgi element is shown in the youngest oogonia which are found near the septal insertion of the ovary, I wish to refer him to the excellent photographs of Foot and Strobell. Plate XLII, photo 29, shows three oocytes, ‘each showing one or more osmiophile granules’.

The Golgi elements in the eggs of Pheretima are in the form of vesicles, each vesicle having an osmiophile rim and a clear central substance. Gatenby (see Gatenby and Nath, 1926) in the fresh eggs of Lumbricus also described a Golgi element as a ‘somewhat irregularly spherical bead’ or a ‘spherule’. He described the rim of the Golgi vesicle as the ‘dictyosome, or thickened edge of the Golgi bead’ which ‘may occasionally be seen as a highly refractive peripheral area’. In fixed preparations (Da Fano), however, Gatenby and Nath described a Golgi element as a sphere, with a thickened edge or dictyosome, which varies in shape from a perfect bananashaped rod to an irregular curved plate. In an earlier section (see Observations) I have shown that the break in the rim of the Golgi vesicle is due either to partial impregnation of the rim, or to the ‘crescents’ with their sphere substance representing optical sections of the vesicles. This can be the only explanation, because in the fresh cover-slip preparations treated with osmic acid for half an hour the rim of the Golgi element is perfectly entire. Even thirty years ago Foot and Strobell studied these Golgi elements in the living eggs of Allolobophora and described them as granules exactly like those which I have shown in my figures of the living eggs of Pheretima.

In face of the clear statement of Gatenby, that in the fresh egg of Lumbricus the Golgi element is a ‘spherule’ or a ‘bead’ with a ‘highly refractive peripheral area’, it is difficult to understand what led Harvey (1929) to make the statement ‘that in the oocytes of Lumbricus, as shown by Nath with Gatenby (1926), the Golgi apparatus can be very easily seen in the living in the form of rodlets or crescents’. I may seem to be labouring this point, but the morphology of the Golgi element in eggs is of fundamental importance in the study of its functions.

Harvey (1929) takes exception to my statement (1928) that the Golgi elements in living eggs can be more easily studied than in the living somatic or male germ-cells because their refractive index is higher on account of the presence of colloids in the form of free fat inside them. In support of his criticism Harvey cites the egg of Lumbricus, and says that ‘in this animal the Golgi apparatus is not concerned with fatty yolkformation, there being no fat in the egg’, although I must add that in 1925 Harvey did mention the existence of fat in this egg. In reply to this criticism I may point out that the Golgi elements of Pheretima can certainly be demonstrated to contain small amounts of fat if the necessary precautions are taken (see Observations), that thirty years ago Foot and Strobell demonstrated the blackening of the Golgi elements of the egg of Allolobophora in five minutes in 1 per cent, osmic acid, and that even in the case of Lumbricus Gatenby and Nath described the Golgi elements as ‘slightly osmiophile’, in spite of the fact that they did not take the necessary precautions. In the case of the oocytes of Culex, also, Nath (1929) has demonstrated the existence of fat in the Golgi vesicles by eliminating to a large extent the decolorizing action of xylol.

Gatenby and Woodger (1920), Ludford (1921), and Brambell (1924) showed that in Patella the Golgi dictyosomes secrete the fatty yolk, although it is very important to note that Ludford clearly mentioned that some of the Golgi elements are directly metamorphosed into such yolk. Brambell (1924) and Nath (1924), however, demonstrated that in Helix and L i t h o b i u s respectively the Golgi element is directly converted into fatty yolk. Nath (1926, &c.), and Nath and collaborators have insisted on the fundamental morphological similarity between a Golgi vesicle and a fatty yolk-vacuole, and have shown in a variety of eggs that the latter is nothing but a swollen Golgi vesicle containing fat. Similarly King (1926) showed that ‘the formation of fatty yolk from the Golgi elements in O n i s c u s is strictly comparable to the process described by Nath in Lithobius’.

Recently, however, Hibbard (1928) and Harvey (1929) have claimed that, in the Amphibian Discoglossus and the crab respectively, fatty yolk arises independently in the cytoplasm. Since I have not personally examined these forms I will neither challenge nor accept these conclusions. Reference may be made to the numerous papers by Nath and his collaborators for the details of the process of the origin of the fatty yolk from the Golgi vesicles, but I will take this opportunity to lay the utmost emphasis that I can command on three points only of great fundamental importance.

1. Hibbard and Harvey appear not to have realized that they are dealing with ‘fat globules’ and ‘fat droplets’ respectively, while I am dealing with a vesicle having a definite membrane containing fat. A simple experiment will explain my meaning and will bring home to them the difference between a fat globule and a vesicle containing fat. If particles of an oil, say clove oil, are thrown into osmic acid they go black immediately. Under the microscope they appear uniformly black and do not show an osmiophilic rim and a lighter central area. If, on the other hand, they place an advanced oocyte (about 2 mm. long) of the cockroach in 2 per cent, osmic acid and examine its contents (after rupturing the egg if necessary), at intervals from about ten minutes to forty-eight hours after immersion in osmic acid they will notice an entirely different phenomenon. After ten minutes’ immersion or even less they will notice that in addition to the solid albuminous yolk-discs (if these have appeared, because there is no definite relationship between the size of the egg and the first appearance of proteid yolk), which appear perfectly white even after many hours’ osmication, there are dark-brownish vesicles of different sizes. Each vesicle shows a definite black rim (due to curvature) and a lighter brownish central area. The smallest vesicles are light brown, while the bigger are black brown, with the intensity of the brown colour increasing according to the size of the vesicles. When the osmication is prolonged, say to half an hour of more, the bigger vesicles look uniformly black without the rim (just like osmicated fat globules), showing that they contain larger quantities of fat, while the smaller vesicles, which are the Golgi elements containing smaller quantities of fat, still show a dark rim and a lighter central area. After still more prolonged osmication, as for example in Kolatschev, all the vesicles look uniformly black. The smaller vesicles can be traced back in fresh osmic acid preparations to the youngest oocyte, where they have a peri-nuclear arrangement, and in which they appear coppercoloured even after forty-eight hours’ immersion. Gradually they spread out in the cell and can be demonstrated as dark vesicles after shorter and shorter periods of immersion because they are becoming more and more fatty. For further details reference may be made to Nath and Piare Mohan (1929). I have mentioned the cockroach only because it is available in all parts of the world, but the same phenomenon may be observed in the spider, in Scolopendrid, Luciola, and D y s d e r c u s, with periods of immersion in osmic acid varying in each case. The difference between the blackening of a fatdroplet in osmic acid and a Golgi vesicle containing fat is illustrated in Text-fig. 17.

In the opinion of the writer the name ‘fatty yolk’ is misleading, because it is not a case of a substance A being metamorphosed into a substance B, but a case of a small vesicle simply enlarging and storing up fat inside it.

2. Let us consider those eggs in which there is admittedly no fatty yolk, namely, the earthworm and the mosquito. Although in these eggs the Golgi vesicles do not swell up, yet they are fatty. This I have demonstrated in Culex (1929) and in Pheretima by eliminating to a large extent the decolorizing action of xylol on the blackened Golgi vesicles, and the same was demonstrated by Foot and Strobell thirty years ago in Allolobophora, and in spite of the fact that they did not take the necessary precautions, Gatenby and Nath (1926) found the Golgi elements of Lumbricus ‘slightly osmiophile’.

3. In Luciola, Nath and Mehta (1927 and 1929) have demonstrated that the Golgi vesicles of even the primordial germ-cells contain fat inasmuch as they are blackened in ten minutes in osmic acid. In Dysdercus, Bhandari and Nath (in press) have shown that the Golgi vesicles are blackened in half an hour by osmic acid even in the earliest oogonia. Now a reference to Harvey’s and Hibbard’s papers will show that their ‘fat-droplets’ or ‘fat-globules’ appear in the cell in the course of oogenesis, while in Luciola and Dysdercus fatty Golgi vesicles (which react to osmic acid in Kolatschev or Mann-Kopsch like the typical Golgi apparatus) are present long before even the oocyte is differentiated.

I hope Harvey and Hibbard will weigh the above conclusions carefully before they challenge the theory of the origin of fatty yolk from the Golgi elements, which I have elaborated after a prolonged study of fresh cover-slip preparations treated with osmic acid, and of the truth of which I am thoroughly convinced. Even if fat arises independently in the Discoglossus egg and that of the crab, as claimed by Hibbard and Harvey respectively, it does not follow that it always does so.