The Cell in Development and Inheritance

1 See Geddes, Sex, in Encyclopcedia Britannua ; Geddes and Thompson, T/ie Evolution of Sex, 1889; Brooks, T/ie Law of Heredity, 1 883; \yatase ('92), The Phenomena of Sex-differentiation. In the greater number of cases the primordial germ-cells arise in a germinal epithelium which, in the cctlenterates (Fig. 72), may be a part of either the ectoderm or entoderm, and. in the higher types, is a modified region of the peritoneal epithelium lining the body-cavity. In such cases the primordial germ-cells may be scarcely distinguish- able at first from the somatic cells of the epithelium. But in other cases the germ-cells may be traced much farther back in the develop- ment, and they or their progenitors may sometimes be identified in the gastrula or blastula stage, or even in the early cleavage-stages. Thus in the worm Sagitta, Hertwig has traced the germ-cells back to

Fig. 72. — Origin of the germ-cells in a hydro-medusa, Cladonevia. [WeismanN.] A. Young stage ; section through wall of manubrium of the medusa ; ova developing in the ectoderm (ec). B. Later stage, showing older ova {o) and "nutritive cells" (ri). The ova contain small nuclei probably derived from engulfed nutritive cells. two primordial germ-cells lying at the apex of the archenteron. In some of the insects they appear still earlier as the products of a large " pole-cell " lying at one end of the segmenting ovum, which divides into two and finally gives rise to two symmetrical groups of germ- cells. Hacker has recently- traced very carefully the origin of the primordial germ-cells in Cyclops from a "stem-cell " (Fig. 74) clearly distinguishable from surrounding cells in the early blastula stage, not only by its size, but also by its large nuclei rich in chromatin, and by its peculiar mode of mitosis, as described beyond.

The most beautiful and remarkable known case of early differenti- ation of the germ-cells is that of Ascaris, where Boveri was able to trace them back continuously thiough all the cleavage-stages to the two-cell stage ! Moreover, from the outset the progenitor of the germ- cells differs from the somatic cells not 07ily in the greater size and rich- ness of chromatin of its nuclei, but also in its mode of mitosis; for in all those blastomeres destined to produce somatic cells a portion of

^^g- 73- — Origin of the primordial germ-cells and casting out of chromatin in the somatic A. Two-cell stage dividing; s. stem-cell, from which arise the germ-cells. />'. Tlie same from the side, later in the second cleavage, showing the two types of mitosis and the casting out of chromatin {c) in the somatic cell, C. Resulting 4-cell stage; the eliminated onromatin at c. D. The third cleavage, repeating the foregoing process in the two upper cells.

the chromatin is cast out into the cytoplasm, where it degenerates, and only in the germ-cells is the sum-total of the chromatin retained. In Ascaris megalocephala univalens the process is as follows (Fig. yi): Each of the first two cells receives two elongated chromosomes. As the ovum prepares for the second cleavage, the two chromosomes reappear in each, but differ in their behaviour ( F'ig. 73, A, B). In one of them, which is destined to produce only somatic cells, the thickened ends of each chromosome are cast off into the cytoplasm and degen- erate. Only the thinner central part is retained and distributed to the daughter-cells, breaking uj) meanwhile into a large number of segments which split lengthwise in the usual manner. In the other cell, which may be called the stem-all (Fig. J I, s), all the chromatin is preserved and the chromosomes do not segment into smaller pieces. The results are plainly apparent in the four-cell stage, the two somatic nuclei, which contain the reduced amount of chro- matin, being small and pale, while those of the two stem-cells are far larger and richer in chromatin (Fig. 73, C). At the ensuing division (Fig. 73, D) the numerous minute segments reappear in the two somatic cells, divide, and are distributed like ordinary chromosomes ; and the same is true of all their descendants thenceforward. The other two cells (containing the large nuclei) exactly repeat the history of the two-cell stage, the two long chromosomes reappearing in each of them, becoming segmented and casting off their ends in one, but remaining intact in the other, which gives rise to two cells with large nuclei as before. This process is repeated five times (Boveri) or si.\ (Zur Strassen), after which the chromatin- elimination ceases, and the two stem-cells or primordial germ-cells thenceforward give rise only to other germ-cells and the entire chromatin is preserved. Through this remarkable process it comes to pass that in this animal only tJic ij^crni-cclls receive the s/n/i- total of the cgg-cJiromatin Jianded dozen from the parent. All of the somatic cells contain only a portion of the original gcrm-substanee.

" The original nuclear constitution of the fertilized (t^^ is transmitted, as if by a law of primogeniture, only to one daughter-cell, and by this again to one, and so on ; while in the other daughter-cells the chromatin in part degenerates, in ])art is transformed, so that all of the descendants of these side-branches receive small reduced nuclei." ^ It would be difficult to overestimate the imi)ortance of this dis- covery ; for although it stands at present an almost isolated case, yet it gives us, as I believe, the key to a true theory of differentiation development,^ and may in the end prove the means of explaining many phenomena that are now among the unsolved riddles of the cell.

Hacker ('95) has shown that the nuclear changes in the stem- cells and primordial eggs of Cyclops show some analogy to those of Ascaris, though no casting out of chromatin occurs. The nuclei are very large and rich in chromatin as compared with the somatic cells, and the number of chromosomes, though not precisely determined, is less than in the somatic cells (Fig. 74). Vom Rath, workin in the same direction, believes that in the salamander also the number of chromosomes in the early progenitors of the germ-cells is one-half that characteristic of the somatic cells. ^ In both these cases, the chromosomes are doubtless bivalent, representing two

A. Young embryo, showing stem-cell {st). D. The stem-cell lias ciiviiled into two. giMut; rise to the primordial germ-cell {g). C. Later stage, in section; tlie iirimordial germ-cell has migrated into the interior and divided into two ; two groups of chromosomes in each. chromosomes joined together. In Ascaris, in Hke manner, each of the two chromosomes of the stem-cell or primordial germ-cells is probably plurivalent, and represents a combination of several units of a lower order which separate during the segmenLution ol the thread when the somatic mitosis occurs.

{a) GnnvtJi and Xutritiou. — Aside from the transformations of the nucleus, which are considered elsewhere, the story of the ova- rian history of the (tg^^ is largely a record of the changes involved in nutrition and the storage of material. As the primordial germ-cells enlarge to form the mother-cells of the eggs, they almost invariably become intimately associated with neighbouring cells which not only support and protect them, but also serve as a means for the elabora- tion of food for the growing egg-cell. One of the simplest arrange- ments is that occurring in coelenterates, where the ^^^'g lies loose either in one of the general layers or in a mass of germinal tissue, and may crawl actively about among the surrounding cells like an Avia:ba. In such cases (hydroids) the (t'g'g may actually feed upon the surrounding cells, taking them bodily into its substance or fusing with them^ and assimilating their substance with its own. In such cases ( Tubiilaria, Hydra) the nuclei of the food-cells long persist in the egg-cytoplasm, forming the so-called '' pseudo-cells," but finally degenerate and are absorbed by the ^ZZ- ^^ would here seem as if a struggle for existence took place among the young ovarian cells, the victorious individuals persisting as the eggs ; and this view is probably applicable also to the more usual case where the ^gg is only indirectly nourished by its brethren.

In most cases, as ovarian development proceeds, a definite associa- tion is established between the Q,gg and the surrounding cells. In one of the most frequent arrangements the ovarian cells form a regular layer ox follicle about the ovum (Figs. 59, 79), and there is very strong reason to believe that the follicle-cells are immediately concerned with the conveyance of nutriment to the o\'um. A num- ber of observers have maintained that the follicle-cells may actually migrate into the interior of the egg, and this seems to be definitely established in the case of the tunicates and mollusks (Fig. 75).^ Such cases are, however, extremely rare; and, as a rule, the material elaborated by the nutritive cells is passed into the o^gg either in solu- tion or in the form of granular or j^rotoplasmic substance.'"^ An interesting case of this kind occurs in the cycads, where, according to Ikeno ('9<S), the egg-cell is connected with the surrounding cells by broad protoplasmic bridges through which cytoplasmic material flows directly into the egg-cell.

Verv curious and su^-scestive conditions occur among the annelids and insects. In the annelids the nutritive cells often do not form a follicle, but in some forms each ej^^g is accompanied by a single nurse-cell, attached to its side, with which it floats free in the body- cavity. In Ophryotrocha, where it has been carefully described by Korschelt, the nurse-cell is at first much larger than the ^^^'g itself, and contains a large, irregular nucleus, rich in chromatin (Fig. 76). The egg-cell rapidly grows, apparently at the expense of the nurse- cell, which becomes reduced to a mere rudiment attached to one side of the ^gg and finally disappears. There can hardly be a doubt, as Korschelt maintains, that the nurse-cell is in some manner con- nected with the elaboration of food for the growing egg-cell ; and the intensely chromatic character of the nucleus is well worthy of note in this connection. Still more in- teresting are the conditions observed by Wheeler ('96, '97) in Myzostoma^ where the young ^g^g is accom- panied by two nurse-cells, one at either end. These cells fuse bodily with the ^Zg, one having " some- thing to do in forming the vacuolated cytoplasm at the animal pole, . . . the other in forming the granu- lar cytoplasm at the vege- tative pole" ('97, p. 42). The polar axis thus deter- mined persists as that of the ripe ovum. This seems one of the clearest

cases showing the establishment of the egg-polarity through the relation of the ^gg to its environment.^ Somewhat similar nurse-cells occur in the insects, where they have been carefully described by Korschelt. The eggs here lie in a series in the ovarian ''egg-tubes" alternating with nutritive cells vari- ously arranged in different cases. In the butterfly ]\vicssa, each ^gg is surrounded by a regular follicular layer of cells, a few of which at one end are differentiated into nurse-cells. These cells are very large and have huge amoeboid nuclei, rich in chromatin (Fig. -J J, A). In the ear-wig, Forficula, the arrangement is still more remarkable, and recalls that occurring in Ophryotrocha. Here each

A. Earlier stage, surrounded by follicle. />. Later stage, showing inward migration and absorption of fol- licle-cells. ^g^ lies in the egg-tube just below a very large nurse-cell, which, when fully developed, has an enormous branching nucleus as shown in Fig. 163. In these two cases, again, the nurse-cell is character- ized by the extraordinary development of its nucleus — a fact which points to an intimate relation between the nucleus and the metabolic activity of the cell.^

In all these cases it is doubtful whether the nurse-cells are sister- cells of the egg which have sacrificed their own development for the sake of their companions, or whether they have had a distinct origin from a verv early period. That the former alternative is possible is shown bv the fact that such a sacrifice occurs in some animals after the eggs have been laid. Thus in the earthworm. Liimbricns tcrrcs- Fig. 76. — Egg and nurse-cell in the annelid, Ophryotrocka, [KORSCHELT.] A. Young stage, the nurse-cell («) larger than the egg {o). B. Growth of the ovum. C. Late stage, the nurse-cell degenerating.

tris, several eggs are laid, but only one develops into an embryo, and the latter devours the undeveloped eggs. A similar process occurs in the marine gasteropods, where the eggs thus sacrificed may undergo certain stages of development before their dissolution. - ib) Diffcnntiation of tJic Cytoplasm and Deposit of Dcntoplasm. — In the very young ovum the cytoplasm is small in amount and free from deutoplasm. As the egg enlarges, the cytoplasm increases enormously, a process which involves both the growth of the pro- toplasm and the formation of passive deutoplasm-bodies suspended in the protoplasmic network. During the growth-period a peculiar body known as the yolk-nucIcus appears in the cytoplasm of many ova, and this is probably concerned in some manner with the growth

of the cytoplasm and the formation of the yolk. Both its qxW\\\ and its physiological role are, however, still involved in doubt. The deutoplasm first appears, while the eggs are still very small, in the form of granules which seem to have at first no con.stant posi- tion with reference to the egg-nucleus, even in the same species. Thus Jordan ('93) states that in the newt {Dicniyctylus) the yolk may be first formed at one side of the ^g^ and afterward spread to other parts, or it may appear in more or less irregular separate patches which finally form an irregular ring about the nucleus, which at this period has an approximately central position. In some Amphibia

Fig- 77- — Ovarian eggs of insects. [Korschklt.J A, Egg of the butterfly, Vanessa, surrounded by its follicle; above, thiee nurse-cells {>i.c.) wiiii branching nuclei; g.v. germinal vesicle. B. Egg of water-beetle, Dyliscus, living; the egg (o.v.) lies between two groups of nutritive cells ; the germinal vesicle sends amoeboid processes into the dark mass of food-granules. the deutoplasm appears near the periphery and advances inward toward the nucleus. More commonly it first appears in a zone surrounding the nucleus (Fig. yS, C, D) and advances thence toward the periphery (trout, Henneguy ; cephalopods, Ussow). In still others {e.g. in myriapods, Balbiani) it appears in irregular jxitches scattered quite irregularly through the ovum (Fig. 78, A). In Braiicliipus the yolk is laid down at the centre of the ^'^^g, while the nucleus lies at the extreme periphery (Brauer). These variations show in general no definite relation to the ultimate arrangement — a fact which proves that the eccentricity of the nucleus and the polarity of the

^^g cannot be explained as the result of a simple mechanical dis- placement of the germinal vesicle by the yolk, as some authors have maintained. The primary ori^nn of the deutoplasm-grains is a question that involves the whole thcorv of cell-action and the relation of nucleus Fig. 78. — Young ovarian eggs, showing yolk-nuclei and deposit of deutoplasm. A. Myriapod {Gcophilits) with single "yolk-nucleus" (perhaps an attraction-sphere) and scat- tered deutoplasm. [Bai.BIAM.]

D. The same with several yolk-nuclei, and " attraction-sphere," s. [Balbiam.] C. Fish {Scorpccna), with deutoplasm forming a ring al)out the nucleus, and an irregular mass of "eliminated chromatin" (? yolk-nucleus). [\'an Hammkkk.] D. Ovarian egg of young duck (three months) surrounded by a follicle, and containing a " yolk- nucleus," j.;/. [Mertens.] and cytoplasm in metabolism. The evidence seems perfectly clear that in many cases the deutoplasm arises in situ in the cytoplasm like the zymogen-granules in gland-cells. But there is now also a very considerable body of evidence indicating that a part of the egg-cytoplasm is directly or indirectly derived from the nucleus through the agency of the yolk-nucleus or otherwise ; and the

subject can best be considered after an account of that bodv. It may be mentioned here, however, that a large number of observers have maintained a giving off of nuclear substance to the cytoplasm, in the form of actual buds from the nucleus (Blochmann, Scharff! Balbiani, etc.) as separate chromatin-rods or portions of the chromatin network (Fol, Blochmann, Van Bambeke, Erlanger, Mertens. Calkins, Nemec, etc.) or as nucleolar substance (Leydig, Balbiani, Will, Lev- dig, Henneguy), but nearly all of these cases demand reexamination.

Fig. 79. — Young ovarian eggs of birds and mammals. [Mertens,] A. Egg of young magpie (eight days), surrounded by the follicle and containing germinal vesicle and " attraction-sphere." B. Primordial egg (oogonium) of new-born cat, dividing. C. Kgg of new-born cat containing " attraction-sphere " {s) and centrosome. D. Of young thrusli sur- rounded by follicle and containing besides the nucleus an attraction-sphere and centrosome (s). and a yolk-nucleus (;'.«.). E. Of young chick containing nucleus, attraction-sphere, and f.itty deutoplasm-spheres (black). F. Egg of new-born child, surrounded by follicle ami containmg nucleus and attraction-sphere.

{c) Yolk-7i?(c!e?(s.—T\\Q term yolk-nucleus or vitelline body ( Dottcr- kern, corps vitcllin) has been applied to various bodies or masses that appear in the cytoplasm of the growing ovarian ^^'^^•^ and it must be said that the word has at present no well-defined mean- ing. As originally described by von Wittich ('45) in the eggs of spiders, and later by Balbiani ('93) in those of certain myriapods. the yolk-nucleus has the form of a single well-defined spheroidal

mass which appears at a very early period and persists throughout the later ovarian history. In other forms there are several so-called ** yolk-nuclei," sometimes of fairly definite form as described in the Amphibia by Jordan ('93) and in some of the myriapods by Balbiani ('93). In some forms the numerous "yolk-nuclei" are irre,e;ular, ill- defined granular masses scattered through the cytoplasm, as described by Stuhlman {^%6) in the eggs of insects. In still others the "yolk- nucleus" or "vitelline body" closely simulates an attraction-sphere, being surrounded by distinct astral radiations and enclosing one or more central granules like centrosomes ( 6"r<y>/'////.v, Balbiani, '93, and Li)uuliis, Munson, '98). Balbiani is thus led to regard the \'olk- nucleus in general as being a metamorphosed attraction-sphere. Miss Foot ('96) has brought forward evidence to show that the polar rings, observed in the eggs of certain leeches and earthworms, are also to be regarded as "yolk-nuclei" (Fig. 102). Henneguy ('93, '96) finally compares the yolk-nucleus to the macronucleus of the Infusoria (!).

In the present state of the subject it is quite impossible to reconcile the discordant accounts that have been given regarding the structure, origin, and fate of the "yolk-nuclei", and from the facts thus far determined we can only conclude that the various forms of " yolk- nuclei " have little more in common than the name. It is, in the first place, doubtful whether the " yolk-nuclei " simulating an attrac- tion-sphere have anything in common with the other forms ; and Mertens ('93), Munson ('98), have shown that the young ovarian ova of various birds and mammals (including man) and of Liinuliis contain one or more "yolk-nuclei" in addition to the " attraction- sphere "(" vitelline body" of Munson). In the second place there seem to be two well-defined modes of origin of the yolk-nucleus. In one type, illustrated by Jordan's observations on the newt ('93), the " yolk-nuclei " arise separately /;/ situ in the cytoplasm without direct relation to the nucleus. The same is true of the small peripheral " yolk-nuclei " of Limulus (Munson). In a second and more frequent type the "yolk-nucleus" first appears very near to or in contact with the nucleus, suggesting that the latter is directly concerned in its formation. The latter is the case, for example, in the eggs of Cyma- too^astcr (Hubbard, '94) .S"j7/<,'-;/c?'/'// /as- (Henneguy, '96), the earthworm (Calkins, '95, Foot, '96), PolyzoiiiuDi and other myriapods (Nemec, '97, Van Bambeke, '98), Limulus (Munson, '98), Cypris (Woltereck, '98), and ]\[olguIa (Crampton, '99). In nearly all of these forms the yolk-nucleus first appears in the form of a cap closely applied to one side of the nucleus (Figs. 80, 81), sometimes so closely united to the latter that it is difficult to trace a boundary between them. At a later period the yolk-nucleus moves away from the nucleus and in

most, if not in all, cases breaks up into smaller and smaller fra^^ments which contribute, directly or indirectly, to the cytoplasmic f^rowth. In all these cases the history of the yolk-nucleus is such as to indi- cate the participation of the nucleus in its formation. Calkins (95) endeavours to show that the yolk-nucleus in Luvibricus is directly derived from the nucleus by a casting out of a portion of the chro- A. Early ovarian egg of Z,?Yr;«(5r/V//^. B. Later stage; fragmentation of yolk-nucleus. C. Ova- rian ^gg of PholcHS. D. Later stage; disintegration of yolk-nucleus. E. Remains of the yolk- nucleus scattered through the cytoplasm. F. Early stage of yolk-nucleus in Mo/^^uhi. G-I. Dis- integration of the yolk-nucleus and enlargement of the products to form deutoplasm-spheres.

matin-reticulum — a result agreeing in principle with earlier obser- vations on other eggs by Balbiani, Henneguy, Leydig, Will, and other observers. This conclusion rests partly on the apparent direct continuity of yolk-nucleus and chromatin, partly on the staining- reactions. Thus when treated with the Biondi-Khrlich mixture (basic methyl-green, acid red fuchsin), the yolk-nucleus at first stains green like the chromatin, while the cytoplasm is red, and this is the case

even after the yolk-nucleus has quite separated from the nuclear membrane. Later, however, as the yolk-nucleus breaks up, it changes its staining power, and stains red like the cytoplasm. The later observations of Miss Foot ( '96) give ground to doubt the conclusion that the yolk-nucleus is here actually metamorphosed chromatin, for by the combined action of lithium carmine and Lyons blue its substance is sharply differentiated from the chromatin. Still later studies by Crampton (99) on Molgula demonstrate that in this case the volk-nucleus is not directly derived from chromatin, but they nevertheless indicate clearly the formation of the yolk-nucleus by or under the immediate influence of the nucleus — a conclusion also reached on less satisfactory evidence by Hubbard, Van Bambeke, Woltereck, and Nemec. The general morphological history of the yolk-nucleus is here closely similar to that of Lumbricus (Fig. 80), except that no direct continuity between it and the nuclear substance was observed. Stained with methyl-green-fuchsin the yolk-nucleus and major part of the nuclear substance stain red, while the scattered nuclear chromatin-granules and the cytoplasm stain green. Millon's test, combined with digestion-experiments and the foregoing staining- reactions, proves that the yolk-nucleus and the red staining nuclear substance consist of albuminous substance and differ widely from the general cytoplasm, which probably consists largely of nucleo- albumins {cf. p. 331). These reactions give strong ground for the conclusion that the substance of the yolk-nucleus, which progressively accumulates just outside the egg-nucleus, is formed through the direct activity of the latter, perhaps arising within the nucleus and passing out into the cytoplasm.

It is possible, further, that 'even the scattered " yolk-nuclei " that seem to be of purely cytoplasmic origin may arise in a similar manner, either, as Crampton suggests, through the early formation and breaking up of a single yolk-nucleus, or in some less Interesting questions are suggested by those " yolk-nuclei," such as occur in GcopJiilus and Limulus, that so closely simulate an attraction-sphere. Munson's observations show that this body ("vitelline body") first appears in the very young ova as a crescent applied to the nucleus precisely as in Molo;iila or Lumbricus, but containing one or more central granules (Fig. 81). In later stages it becomes spherical, moves away from the nucleus, and assumes the form of a typical radial attraction-sphere with concentric microsome- circles and astral rays. It is hardly possible to doubt that this body in Limulus is of the same general nature as the yolk-nucleus of Lumbricus, Molgula, Cypris, Cymatogaster, or PJiolcus ; and if it be a true attraction-sphere in the one case we must probably so regard it in all. This identification is, however, by no means complete;

and even Munson's careful studies do not seem definitely to establish its connection with the attraction-sphere or centrosomc of the last oogonium-division. That a body simulating an attraction-sj)here and containing a central granule may arise dc ?iovo in the cytoplasm is shown by Lenhossek's observations on the spermatids of the rat (p. 170); and the central granule is in this case certainly not a centrosome, since the true centrosomcs are found in another part of the cell. It is quite possible that the "vitelline body" of Liviiihis may have a similar origin. Nemec ('97) finds in Polyzoniuvi in the earliest stages a single body applied to the nucleus and later two bodies, one of which enlarges to form a cap-shaped yolk-

Fig. 81. — Forms of yolk-nuclei in Limidus and Polyzonium. [./-C, MrNSON; D-F, Np:mkc.] A. Very young ovarian eggs oi Llmulus ; at the left "vitelline body" (t-) in the form of a cap on the nucleus; at the right older egg showing astral formation. B. Older stage of the same; "vitelline body" in the form of an attraction-sphere with central granule. C. Peripheral "yolk- nuclei" {y.n.) in Liviulus. D. Very early ovarian egg of a myriapod. Polyzonium, with yolk- nucleus. E. O der egg with yolk-nucleus and astral body ^a). F. Still later stage, beginning disintegration of the yolk-nucleus.