The Spermatogenesis of Anasa Tristis ...

Front matter

Google to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books

Introduction

The following paper contains the results of a study of the male germ cells of Anasa^ one of the Heteroptera, corroborating in the main an earlier work by Henking on Pyrrochoris^ although a number of new points are brought out to supplement the excellent work of my predecessor.^

Henking ('91) described the structure of the spermatogones and the preparation of the chromatin for the spermatocyte divisionSy and though many of the details escaped him, he certainly, it seems, made a correct estimate of the value of these divisions. His description of the formation of the spermatid is especially good, and with the exception of what are probably merely specific variations, the process in Anasa approximates very closely to what he has described. Henking was also the first to describe the peculiar chromatic body in the spermatocytes to which particular attention has been paid in this paper under the name of the small chromosome.

Wilcox ('95, '96), the only author besides Henking who has followed the entire history of the spermatozoa in the insects, found that in CcUoptenus^ one of the Orthoptera, the process varies in many essentials from that in the Heteroptera, the most fundamental difference lying in the fact that he finds no longitudinal splitting in the spermatocytes ; and although tetrads are formed (by conjugation), both subsequent divisions are transverse or " reducing " ones.

Besides these, several authors have given a partial account of the history of the spermatozoa. Perhaps the most important of these works is Vom Rath's ("92) on Gryllotalpa^ where the chromatin changes, and especially the process of reduction,

^ A coariderabte number of observations have been made on other species of Hemiptera-Heteroptera, the results of some of which have been akeady described (Paulmier, '98). Among the forms examined have been spedes of Euchistus, Brochymenay Mortnideoy MurgantiOy Coenusj Podisus^ Mineus^ and Nttara of the family Pentatomidae ; CharUsterus^ Euthcctha^ and Alydus of the family Coreidae, and Myodocka and Lygaeus of the family Lygaeidae. From these observations it appears that there may be a very great difference in details between the various genera (cf, difference in tetrad formation described before, '98), but the main restdts are the same in all cases. Some observations on the spennatogenesiB of Papilio gave interesting restdts which will be mentioned in the text were first carefully worked out. Platner ('89), on the other hand, confined himself to a study of the achromatic structures in the Lepidoptera.

Montgomery in three papers (•97-'99) has treated of the spermatogenesis of Pentatama (Euckistus), one of the Heteroptera, carrying it, however, only up to the formation of the spermatid. His results di£Fer in many respects from mine ; in regard to the reduction question his conclusion is that while a longitudinal splitting of the chromosomes and consequently an equation division usually occurs in the spermatocyte divisions, both divisions may in some cases be transverse, and therefore both reducing. He gives an interesting account of the body which he calls the '' chromatin nucleolus," and which I have described under the name of the small chromosome.

In the present work I have endeavored to follow the entire history of the male germ cells, from the spermatogones to the mature spermatozoa. Among the more important results may be mentioned the probable continuity of the centrosome throughout all stages, of interest on account of its bearing on the discussion as to whether or not the centrosome is a permanent organ of the cell ; the occurrence of a longitudinal splitting of the chromosomes in the prophases of the first spermatocyte division and its relation to the formation of tetrads, a question of considerable interest in view of several works on insect spermatogenesis, and in which my observations point to an agreement with the copepod type of tetrad formation ; the history of a chromatic body which has been described recently by several authors, and a suggestion as to its meaning, which is of interest as bearing on the theory of the individuality of the chromosomes ; and a partial explanation of the giant spermatozoa.

I wish to express my best thanks to Professor Wilson for •his advice and encouragement throughout the progress of this work. My thanks are also due to Dr. Calkins for his kindness in reading the manuscript of this paper.

I. Material and Methods.

The insect on which most of the following observations were made is Anasa tristis^ De Geer, of the family Coreina, order Hemiptera-Heteroptera. It is commonly known as the " squash bug," its favorite food plant being the squash, though it also feeds on the pumpkin and probably other plants of the same family. The adult bugs winter under the bark of trees and under stones or in crevices, and appear during the latter part of June and July, the eggs being laid during the latter month. The larval period lasts about four weeks, and abundant material for spermatogenesis can be obtained in any desired stage. Most of the material used was obtained from specimens of the two instars preceding the last moult, the majority of the cells of the testis then being in the spermatocyte division or early spermatid stage. The later growth of the spermatozoa takes place after the last moult and during the winter.

The testes were dissected out in normal salt solution and brought as rapidly as possible into the fixing fluid. For fixing, a number of agents were tried, but the best results were obtained with Hermann's and Flemming's fluids, used for fifteen to thirty minutes and followed by thorough washing in running water. The material was then gradually transferred through successive alcohols to 90 per cent, in which it was kept It was cleared in clove oil and imbedded in paraffine, the immersion in the heated paraffine being made as short as possible, usually not more than fifteen minutes. Sections about 5 aa thick were then cut on a Minot microtome.

For staining, Heidenhain's iron haematoxylin was employed almost exclusively. Other stains were tried, but none gave such satisfactory results.

II. Description of Testis.

The paired testes lie on the ventral side of the abdominal cavity, one on each side of the alimentary canal. Each adult testis is a flattened pear-shaped body, consisting of seven approximately equal lobes arranged in a single row and with no indication of a division into groups, as in Pyrrochoris (Mayer, '74), or of a difference in size, as in Pentatoma (Montgomery, '98). During the first few instars the testis is white, but after the third moult a faint yellow color makes its appearance toward the free end. This gradually deepens to orange, and then just before the fifth or last moult, or in some cases earlier, — different individuals varying in this respect, — a red pigment appears between the lobes and spreads out in streaks across them. By the time the insect emerges from its hibernation the red pigment has extended over the entire testis, which is thus of a uniform red color. The pigment does not extend over the vas deferens.

Surrounding the entire testis and continuous with the vas deferens and forming a partition between the lobes is a layer of connective tissue (PI. XIII, Fig. i, a). In it are trachea (/.) and numerous somatic nuclei which divide by mitosis. In this layer also are the red pigment granules which give the characteristic color to the adult testis. Besides this layer, there is another thin layer of connective tissue lining each one of the lobes (PI. XIII, Fig. i, b) and spreading across the lobe in thin sheets to form partitions between the cysts (PI. XIII, Fig. i, V), It also forms a considerable mass at the lower end of each lobe, closing, in the immature testis, the opening into the vas deferens. At the points where the cross walls are given off are frequently found nuclei, the interesting feature of which is that they may be of several times the size of ordinary somatic nuclei (compare in PI. XIII, Fig. i, d and d^ with n, a somatic nucleus, or with ^, a spermatocyte of the stage of Fig. 22) and contain a correspondingly large number of chromosomes. They have also an extra number of centrosomes, as many as eight having been found. Such cells are possibly due to the fact that after an ordinary normal division of a somatic nucleus the two groups of chromosomes fail to separate very far (perhaps on account of the limited space) and both become surrounded by a single nuclear membrane, on which both centrosomes come to lie. This process may be repeated several times.

The reproductive cells of the testis are arranged in cysts surrounded by the connective-tissue membrane above mentioned. There are generally two longitudinal rows of cysts in each lobe, making fourteen rows in the testis. All the cells of a cyst are in the same stage of development. The young testis is entirely filled with spermatogones (adopting the terminology of St. George), but as it grows older the spermatogones at the lower end change successively to spermatocytes and then to spermatids. Thus by the time that the insect is ready to undergo its last moult practically all stages can be found. Thi^ spermatogones, both isolated and in cysts, are at the extreme free end of the testis, forming the Keimzone of O. Hertwig. Then follow, with limits fairly well defined, cysts of growing spermatocytes (Wachtumszone), dividing spermatocytes (Reifzone), and finally developing spermatids. Successive cysts, however, do not usually contain successive stages, there being frequently a considerable gap between. This is probably to be explained by their mode of development ; a cyst of spermatogones changing to spermatocytes which undergo considerable development before the following cyst of spermatogones is ready to go through the same changes. Frequently, however, several cysts may run through the same changes together.

III. The Spermatogones.

The extreme free end of the testis is occupied by a number of spermatogones, either single or in small groups not yet surrounded by a membrane. The number of cells in these groups increases by division, and when a group has reached a certain size, it becomes surrounded by a connective-tissue wall continuous with the lining of the follicle (PL XIII, Fig. i, b). Thus is formed a spermatocyst within which the cells gradually assume a conical shape with their apices turned toward the center (PI. XIII, Fig. 8).

The structure of the cytoplasm is difficult to make out in these cells, but it appears to be reticular rather than alveolar. At the tip of the resting cell is a structureless mass (PI. XIII, Fig. 2, e) which is formed from the remains of the intermediate spindle fibers of the preceding division and is therefore homologous with the " Zell-Koppel " of Helix (Zimmerman, '91). Its mode of origin will be described later. Another body of similar appearance (PI. XIII, Fig. 2,/) is found on the nuclear membrane on the side towards the apex of the cell. In it, lying close against the membrane, is a minute body which I believe to be a centrosome. The chromatin in small granules is arranged in a fine network scattered evenly through the entire nuclear space. Besides this network there are two hazy, indefinite masses staining with chromatin stains (PL XIII, Fig. 2, a), and a nucleolus staining with plasma stains.

In preparation for division the chromatin granules become coarser (PI. XIII, Fig. 3) and arrange themselves in a number of threads (segmented spireme) which shorten, become thicker, and then split longitudinally (PI. XIII, Fig. 4). No single spireme thread was found. These threads soon divide transversely into a definite number of shorter segments, which gradually become more compact (PL XIII, Figs. 5 and 6).

While this has been going on, the centrosome appears to divide, for two small, deeply staining particles can frequently be seen within the hazy mass ; these then separate more and more, stages like PI. XIII, Fig. 5, showing them at varying distances from each other, being frequently found. Finally they arrive at opposite sides of the nucleus. The hazy mass surrounding the resting centrosome disappears, but until the daughter-centrosomes are at opposite sides of the nucleus there is no trace of astral rays. Thus it seems probable that this mass takes no part in the formation of the asters and cannot be regarded as an attraction sphere. It perhaps corresponds to the body which Meves f 96) describes under the name of " idiozom." As soon, however, as the centrosomes have attained their position at opposite sides of the nucleus a few astral rays appear. Then the nuclear membrane disappears and spindle fibers are formed between the centrosomes and chromosomes, probably by a metamorphosis of the linin network of the nucleus. There is in Anasa only a single, rather thickened fiber from each centrosome to each chromosome. The nucleolus disappears with the nuclear membrane.

During these changes the chromosomes become more compact and assume a regular rectangular outline (PL XIII, Fig. 6)} The former longitudinal split is clearly indicated by a slight constriction at the ends of the chromosome (PL XIII, Fig. 6).

They are then drawii into an equatorial plate, which is at right angles to the long axis of the cells, and their longitudinal split is in the plane of division (PL XIII, Fig. 8).

A cross-section of the nuclear plate (PL XIII, Fig. 9) shows

^ In the period represented by PL XIII, Figs. 7 and 9, adjoining chromosomes, apparently without regard to their origin from a common segment of the spireme, may become connected by chromatin bands (see figures).

that there are twenty-two chromosomes, and that they are not all of the same size. The limits of ordinary variation are shown approximately in a and ^, PL XIII, Fig. 9, but there is nothing constant about this variation. Two of the chromosomes^ however^ are invariably found to be very much smaller than the smallest of the others^ though in all other respects similar to them and connected with them by chromatin bands. They are represented by c and d, PI. XIII, Figs. 6 and 9. It has not been possible to determine certainly their origin, but it appears probable that they are formed from the two masses of chromatin described in the resting nucleus and figured in a^ PI. XIII, Fig. 2. Thus it would follow that they retain their identity throughout the whole spermatogone period, and, as will be described later, throughout the spermatocytes also, and form what I have called the small chromosomes.

The chromosomes, including the two small ones, then separate across the equatorial plane (the region of the preceding longitudinal split), and the two daughter-groups are drawn towards the centrosomes, leaving between them a considerable bundle of intermediate fibers (PI. XIII, Fig. 10). Arrived at their final position, the chromosomes separate, their outlines become irregular, and the groups become surrounded by a nuclear membrane (PI. XIII, Fig. 11). This nucleus becomes round, the chromosomes break down more and more, and finally give rise to a reticulum, as shown in PI. XIII, Fig. 2.

As a group of chromosomes approaches the end of its movement the cell loses its conical shape and becomes more cylindrical, and the Zell-Koppel of the preceding generation is cast off, remaining for a time as a small isolated mass which ultimately disappears. The bundle of intermediate fibers, having broken away from the chromosomes, constricts in the middle, where a row of granules appears, forming a cell plate. A constriction appears around the middle of the cell and this gradually deepens until the cell becomes divided in the plane of the cell plate. The round cells are now scattered irregularly in the cyst, but soon the side of the cell containing the remains of the spindle fibers commences to elongate in the direction of the center of the cyst, and the cells assume the characteristic arrangement of the spermatogones. The spindle fibers lose their fibrillar appearance and become a homogeneous mass.

These divisions continue until the cyst reaches a certain size, containing apparently about two hundred and fifty cells. These cells then become spermatocytes.

Food Cells. — In the region where the spermatogones are beginning to transform into spermatocytes, all the cells of certain cysts undergo a process of degeneration, probably to form food cells. The chromatin first collects around the nuclear membrane in lens-shaped masses (PI. XIII, Fig. 13), staining intensely with haematoxylin, and these then fuse into one large round mass, though several smaller ones are often formed (PI. XIII, Fig. 14). The cytoplasm also undergoes degeneration, apparently losing its reticular character and becoming homogeneous. Later the entire cyst disappears, leaving no trace. Nothing similar to such cells is found in any other region of the testis, and this disappearance at the point of greatest growth makes it seem probable that these cells are used as food material for the adjoining cysts of spermatocytes.

The occurrence of such degenerating cysts is somewhat variable ; in some testes they are present in considerable numbers, forming a distinct row separating the spermatogones and spermatocytes, while in others almost none are to be found. It is possible that this is due to a difference in nutrition, the greater number being found in the testes of poorly nourished individuals. If this interpretation of their function is correct, they would be analogous to the " nurse cells " described by Korschelt and others in the ovaries of insects. Similar structures have been described in Ascaris (Hertwig, '90), and in Pyrrocharts (Henking, '92), where they are regarded as degenerating cells. In Cicada and Caloptenusy also, similar structm-es are known (Wilcox, '97), but no fimction has been attributed to them.

IV. The Maturation Divisions.

I . Origin of the Tetrads,

a. Early History of the Chromatin, — After the last spermatogone division the round cells, with their cytoplasm reduced to a very small amount, are scattered irregularly within the cyst wall. They show no tendency to become pointed or to assume the characteristic arrangement of the spermatogones, but as the cells increase in size they become polygonal by mutual pressure. The remains of the spindle fibers have entirely disappeared.

The chromosomes become separated and spread out over the nucleus in the same way as in the spermatogones, after which they break down into a confused mass of filaments (PI. XIII, Figs. 15 and 16), which elongate, become thinner, and show a tendency to collect at one side of the nuclear vacuole. This stage of extreme tenuity does not last long, for the filaments immediately begin to contract and thicken, until finally they reach the condition shown in PI. XIII, Fig. 17, where a number of short rods, staining deeply with haematoxylin, are collected in a little over one-half of the nuclear vacuole. Another phenomenon peculiar to this period is the disappearance of the nuclear membrane. The result of the foregoing changes appears to correspond to the " synapsis '* stage of Moore (*95). I am convinced that it is a normal occurrence which regularly takes place in the development of the chromatin segments, and not, as some have maintained, an artefact due to the action of the fixing agent on the chromatin while in a certain condition. It is of general occurrence at this particular region of the testis, whatever fixing agent has been used, and a series of stages can always be found leading from the metaphases of the spermatogone division to this stage, and from it to the longitudinally split segment of the later period.

Moore C^S), who first described a similar process in Elasmohranchs^ gave to it the name synaptic phase. His description applies pretty closely to the process in Anasa, there being first an increasing fineness of the chromatin reticulum which segregates at one side of the nuclear vacuole, and from which arises the segmented spireme. Brauer (*93) found that in

I

THE SPERMATOGENESIS OF AI^ASA TRISTIS, 233

Ascaris also, in the stage preceding the first spermatocyte division, the chromatin became collected at one side of the nuclear vacuole and lost almost all traces of its thread-like character. Following this is the segmented spireme. Miss Sargant, in her last paper on Lilium ('97), came to the conclusion that the S3niapsis is a normal occurrence and not an artefact, since she found it in the living cells. The three criteria which she gives of the synaptic phase in plants will, however, not altogether apply to the process in Anasa^ for, while the contraction of the chromatin (the most characteristic phenomenon of the stage) and the disappearance of the nuclear membrane take place, there is no trace of any solution of the chromatin, and no nucleolus is present at this time.

The synaptic phase is of widespread occurrence at this period of the chromatin development, though it appears to vary considerably in the extent to which it is carried ; in Lilium^ for instance, apparently involving a partial solution of the chromatin» while in Anasa this does not take place. No explanation of these occurrences has, as far as I am aware, been attempted ; but it seems, in Hemiptera certainly, to be a part of the process by which the " pseudo-reduction '* of the chromatin takes place.

b. Formation of the Tetrads, — The short, thick rods of the synapsis stage (PI. XIII, Fig. 17) elongate somewhat, separating into granules, and each of these rods or segments splits longitudinally (PI. XIII, Fig. 18). Thus we have arrived at the condition of a longitudinally split, segmented spireme which has not gone through any single spireme. It is still impossible to count the number of these segments, but from their later development it will be seen that there are ten, i,e.^ one-half of the number in the somatic cells^ if we neglect the two small chromosomes which will be accounted for later. There has, therefore, been a reduction in the number of segments without a division, ^^., a pseudo-reduction (RUckert). These short longitudinally split segments immediately elongate, the granules of which they are composed becoming smaller and moving farther apart (PI. XIII, Fig. 19). The two rows then separate in the middle and twist around each other (PI. XIII, Figs. 20 and 21). The two halves of each segment usually remain connected at their ends to form a ring, but in some cases the separation may be complete, though the two halves always remain close together. In this stage the segments resemble those figured by Hacker ('92) in the copepods. Immediately after, and perhaps also during this process, the distinct rows of granules of which the segments have been composed become broken up and fused together into larger, diffused, hazy masses, showing less affinity for the stains. This change is accompanied by a certain amount of growth, or at least expansion of the chromatin.

The segments now alter their shape through the disappearance of the twist, and a bend appears at the middle point of each half, so that the segment has the shape of two V% joined together by the extremities of their legs (PI. XIII, Figs. 22 and 23, and Text-fig. i). The apices of the F's (Text-fig. i, d) represent the ends of the original segments, and the space between the legs the longitudinal split. The extremities of the legs of the F*s (Text-fig. i, b) represent the middle point of the segments and correspond to a transverse division.

These V^% may either extend out in the same plane (Text-fig, i), or they may be bent toward each other (Text-fig. 3).^

From these are formed the tetrads by a condensation of the chromatin, the bent ones becoming straightened out during the process (Text-figs. 2-4). The segments usually retain their double P^shape for some time, though considerable variations may occur, — such, for instance, as is represented in Text-fig. 5* where the limbs of the V% are closely pressed together and the transverse division is represented only by a swelling of the middle of the segments. More rarely it is the longitudinal division which is obscured.

In the great majority of cases in all stages the segment is longer in one direction than in the other, the longer axis representing the earlier longitudinal split. In some cases, how-

^ An interesting and perhaps suggestive point which I have mentioned before (Paulmier, '98) is that the apices of the r's may be brought so close together as to give the impression that the body is formed by two longitudinal divisions instead of one longitudinal and one transverse. It is possible that some clue to an explanation of the difference between the two types as shown in Ascaris and in the copepods may be sought in some such process as this. This arrangement of the chromatin in the form of two F's is apparently typical of Anasa and perhaps its nearest allies, and is not found in any Lygaeidae or Pentatomidae yet examined.

ever, the legs of the F*s are drawn so far apart as to form a square and it is impossible to tell which is the longitudinal and which the transverse division. It is possible that in some cases they may be drawn so far apart that the transverse axis may become the long axis of the segment, though if this does occur it is certainly exceptional. The only other circumstance regarding which any doubt may arise as to the position of the two divisions is where the two P^'s are bent close together. In such cases it is frequently impossible to tell which angles represent the apices of the V\ and which the extremities of the legs, and a reversal of the axes may take place as in the preceding case.

The result of this condensation is a tetrad body as represented in PI. XIII, Fig. 25 and Text-fig. 6, with one of its axes

Fio. I. Fio. s. Fio. 3. Fio. 4. FxG. s* Fw*- ^

longer than the other. As the body has been followed through all its transformations it is possible to say that the long axis of this stage represents {for the great majority of cases at least) the long axis of the early stage and therefore corresponds with the longitudinal split.

c. The Small Chromosomes, — The history of the two small chromosomes of the spermatogone divisions may be here described. If these bodies are chromosomes, traces of them should be found in the stages of spermatocyte growth, and I am convinced that the body about to be described corresponds with them. When the chromosomes are breaking down to form the reticulum (PI. XIII, Fig. 15), a portion of the chromatin remains unaltered at one side of the nucleus in a single round mass, which persists throughout the synapsis stages (PI. XIII, Figs. 16 and 17, x). In the stage of the segmented spireme (PI. XIII, Fig. 19), it is very plainly seen as a round body, staining much darker than the chromosomes and lying at one side of the mass of chromatin. When the chromatin fills the nuclear vacuole, this body lies against the membrane — a position which it retains throughout all its further changes. During the stages shown in PI. XIII, Figs. 19-21, it remains as an irregular mass, which, on account of its intense staining power, shows very clearly against the fainter chromatin. In PL XIII, Fig. 22, it is seen to have elongated and to have become split in the direction of its long axis — a process which I regard as analogous to the longitudinal split of the ordinary chromosomes. In PI. XIII, Fig. 23, a second or transverse division has appeared, and thus the body has given rise to a tetrad precisely homologous with the larger tetrads, though it has omitted many of the stages which the latter have gone through. The fact that there is a single one of these bodies, while there were two in the spermatogones, is probably due to the pseudo-reduction, the two being here connected together. This small tetrsid remains close to the nuclear membrane, and together with the other chromosomes is drawn into the equatorial plate of the first spermatocyte division.

d. The Achromatic Structures. — During the growth of the nucleus the cytoplasm undergoes a corresponding increase in size. The cells retain for a considerable time their polygonal shape, but as the nucleus prepares for division the angles become rounded and the cells separate from each other. In the cytoplasm around the nucleus there is formed a considerable mass of large granules of a rather indefinite form and staining faintly with the plasma stains (PI. XIII, Fig. 22). These have been observed by Henking ('91), who gave to them the name " Dotter-Massen" (yolk-masses), regarding them as homologous with the yolk of the egg, since they are laid down at a corresponding period in development.

A number of bodies have been described in the male germ cells of this stage, but it appears doubtful whether they are homologous with these. Erlanger ('97) describes, under the name '* Centrodeutoplasm,** granules which pass through the cell during division and collect around the centrosome in the resting stage. Meves ('97) applies to corresponding bodies the term

" Idiozom,'* and states that they disappear during division. Montgomery (*98) uses the term " Idiozom '' for a body which appears to have characteristics very similar to the yolk granules of Anasa, In later spermatocyte stages it is scattered throughout the cytoplasm, and though some of it disappears during division the rest persists into the spermatid. Montgomery claims that the idiozom has some connection with the centrosomes, but from the study of several genera of the Pentatomidae {Euchistis^ Brochymena) I think there is some reason to doubt the correctness of his conclusions here and believe that the centrosome, as in Anasay moves around on the nuclear membrane entirely free of any idiozom or yolk granules. Yolk granules surrounded by a clear space such as Montgomery describes are not to be found in Anasa,