WHI

By DR. J. A. OSBORNE.

On the 4th of August last, I received by post from Mr. W. R. Jeffrey, of Ashford, Kent, a small batch of seven of these eggs which had been laid on glass some time on the morning of the 2nd. The batch had a somewhat greasy whitish appearance, and on closer inspection the eggs were seen to be arranged in an oval or ring of six surrounding a central one, but all overlapping, in such a manner that the central egg was overlain by three at one end of the oval, and overlay the three at the other end; whilst of the former three the middle (or remotest) egg overlapped the other two, and of the latter three the middle (or nearest) egg was overlapped by them. The lateral pair of the first three also overlapped the lateral pair of the second three, but the eggs of each pair did not touch each other. In this arrangement there was one egg (only) at the nearest extremity of the long axis of the group, which lay directly on the glass without overlying a part of any other; and one (only) at the other extremity of the axis, which, overlying two others in part, was itself not overlapped

by any other. The conclusion appears inevitable that these eggs were respectively the first and last laid in the group: otherwise a later egg must have been partially inserted beneath one already deposited on the glass-a supposition which the character of the eggs themselves would appear to negative decidedly. I found it convenient, taking the longer axis of the group as a meridian running north and south, and designating the central egg as "C," to distinguish the others by the points of the compass, as N., S., N. W., &c., and regarding the group always from the free side. The moth, then, in depositing her eggs, must have proceeded along the glass in a general direction from south to north; and the eggs must have been laid in the following order :-S., S.W. and S. E., C., N.W., and N.E., N. The order of precedence in the lateral pairs is not determinable from these premises. Subsequently Mr. Jeffrey sent me the shells of another group of nine of these eggs, laid at the same time, which give some further insight into the method of oviposition. In this group, which has a transversely elongated rhomboidal form, the extreme lateral eggs, E. and W., which were absent in the first, are present. Here it is plain that the eggs were laid in rows of three each, forming an acute angle with the meridian, and each row beginning a step further east or west and further north than the row before it. The order must (most likely) have been S., S. W., W.; S.E., C., N.W.; E., N.E., N.; or else, S., S. E., E.; S. W., C., N. E.; W., N. W., N. I enclose diagrams (Figs. 30 and 31), the better to illustrate my meaning, but if it should be inconvenient to engrave these, the grouping may be very well imitated with the requisite number of pence or half-pence, taking the vertical line of the figure on the coin to represent the long axis of the egg. In all cases the long axis of the egg-oval lay north and south, i.e. parallel with the meridian of the group; and, as subsequent observation showed, in

WHILST according the first place in embryological research to the method of investigation by means of sections, Dr. August Weismann, in his latest work ("Beitr. z. Kenntniss d. ersten Entwickelungsvorgänge im Insektenei," Bonn, 1882), is yet of opinion that the older method, by "continuous observation of the living and developing egg," has of late years been much underrated. "There are," he says (loc. cit. p. 2), "certainly phenomena of development, where the method by section fails us altogether, and of whose course, nay, very existence, only direct observation gives us any intelligence." Perhaps there are few eggs of insects which, owing to their extreme flatness and transparency, are better suited for direct observation during development than those of Botys hyalinalis. They are small oval discs of about 2 3'-2.5 mm. in length by 2 mm. or rather more in breadth, not thicker in proportion than the body of a sole or plaice is to its diameter, and thinning off in like manner to a sharp edge at the circumference. The shell is transparent as glass, and the view but little impeded by the somewhat coarse reticulations of the chorion in irregularly polygonal fields with linear borders and uneven areas.

all cases the head of the embryo lay south. The question now arises: does the south end of the egg correspond with its lower pole or first laid end? This question must be answered affirmatively, unless we are prepared to admit either, on the one hand that the eggs in laying could have been partly slid under ones already laid, which their thinness and delicacy and their firm adhesion to the glass and to one another seems to render impossible; or, on the other hand, that the ovipositor of the moth, whilst

lower pole, we arrive at the conclusion (somewhat important, inasmuch as it is at variance with the statement of Leuckart that "the upper pole in all cases contains the head end of the embryo."-See Entom. Month. Mag. vol. xx. p. 146), that in these eggs the head of the embryo normally occupies the lower pole of the egg. The physiological reason is obvious. In those cases where (as in Pieris brassica) the egg is attached by its lower pole to the food-plant, the escaping larva, if its head occupied that end,

depositing the egg is bent in under her venter so as to extrude the egg with its lower pole to the front. Such an inversion of the normal orientation does take place in the case of the sawfly (Zaraa fasciata), and probably also, more or less, in the case of other insects laying their eggs in mines or in the ground, or in other situations where a long ovipositor is required, but I am ignorant that there are any grounds for such an assumption in the case of Botys. Taking it for granted then that the south end of the egg is its

Fig. 38.

Fig, 39.

would not only have to eat its way out of the shell, but also through the substance to which the egg was attached; while in the case of Botys if the head of the larva occupied the upper pole, it would, after eating its way through its own shell, come in contact with the egg above it with disadvantage to itself and probable destruction to its neighbour.

When I received the eggs on the 4th of August, they were already in the third day of incubation, and presented the following appearances. The yolk

granules were aggregated in spherular masses, which again were arranged in two groups: an annular larger mass having a clear space outside between it and the shell, and a clear central area in which the smaller mass lay in an irregularly curved spiral or crescentic form. For the sake of convenience, I shall call these two masses of [opaque yolk the Annulus and the Spiral. The former was composed of yolkspherules (i.e. aggregated masses of yolk-granules), arranged, somewhat loosely at first, but by degrees more compactly, so as to flatten their sides and give them a polygonal form, in a flat oval ring, having the same contour as the egg itself, but gradually wider at one end than the other, not unlike a horse-collar. The broader end was at first in all cases (except, perhaps, N.W.) towards the south end of the egg.* It maintained its shape with little change, except some diminution in width and increase in density, all the time that the eggs were under observation, i.e. till the 9th August, when they were unfortunately destroyed by an accident a day or two before hatching would have taken place. The spiral, on the contrary, was constantly undergoing change, and it soon became evident that its convex border was the true seat of development. Regarded as a mass of unchanged yolk, it had, on the 5th August, somewhat the appearance of a double scroll (Fig. 32), something like the capital of an Ionic pillar, and with a stalk between roughly corresponding to the shaft of the pillar, or, at least, the upper end of it. The volutes of this scroll or double spiral were not similar. One was sharp, dark and well defined; the other vague and changing, and made up of looser granules of yolk. In all cases the sharper welldefined volute lay towards the north. The stalk or shaft between the volutes pointed in a general direction east or west, and after some hours became more slender and formed an attachment with the annulus on the inside. At this time the scroll was not unlike the vertebrate embryo in an early stage with its umbilical stalk, and, as no doubt yolk granules passed into the embryo by this channel, there may have been something functional in the resemblance as well.

To explain this annular arrangement of the foodyolk it is necessary to refer to the formation of the amnion. After the yolk has become surrounded by the growth of cells called the blastoderm, and after the germinal stripe, or foundation of the embryo, has been differentiated along one side of this blastoderm, a double fold of the latter grows up all round the circumference of the germinal stripe and finally closes in over it, the edges of the fold fusing together and the two layers (of blastoderm) of which it is composed at the same time separating from one

Under date August 7th, I have the following note. "In the three northern eggs it is the northern half which is widest now-in the three southern, the southern half, and more distinctly so."

another. The inner of these, continuous with the embryo itself, and lying immediately over it, is the amnion; the outer, continuous with the blastoderm surrounding the yolk, is the serous membrane. Two sacs are thus formed, the one within the other, and between them lies the yolk. In the lepidopterous egg the yolk next finds its way into the space between the amnion and the serous membrane, flowing over the former, and depressing it and the embryo beneath it till both are completely submerged in yolk and consequently hidden from view. But owing to the extreme flatness of the Botys' eggs, little or no yolk finds its way to the sides of the embryo, but is constrained to lie in a ring around it, leaving the centre clear, except that part immediately beneath the germinal stripe, which by the involution of the two extremities of the embryo becomes soon reduced to a narrow capitate peduncle, the "umbilical stalk," &c. (Fig. 33).

A glance at the diagrams (Figs. 30 and 31), will show that the umbilicus did not lie uniformly in one direction east or west. When it had its attachment to the annulus on the west, this meant that the venter of the embryo was facing east, and that (having its head south) it was lying on its right side or with its right side next the glass: and vice versa. This arrangement may be very well imitated with the half-pence by placing them upside-down, and with the queen's head up to represent the embryo on its right side, and the figure of Britannia up to represent it lying on its left side with the venter looking west. The only thing to be noted in this connection (and the small number of the total observations makes it the less reliable), is that in both groups of eggs all the lateral eggs, i.e. those not situated on the central axis or meridian of the group, had their umbilici directed towards the meridian, and their ventres consequently looking. outwards. To this statement there were two exceptions in the group of nine; viz. N. W., which did not develop at all, and N.E., which had its venter looking inwards, but failed to hatch out.

There was a considerable difference in the rate of development of the different eggs in the group of seven. S.W. appeared to be the most advanced, although latterly N.E. was not much if anything behind it. But N.W. was all along several hours behind the others in its development. These might be arranged in the order of development thus :-S., S.E., N., C.; but with smaller differences between them. On the evening of the 6th August the umbilical stalk and spiral had dwindled to a small dense inverted cone with a rounded base and having its apex at the annulus (v. Fig. 33). On its north side a sort of notch separated the sharp curved bird-like beak from the rest of the stalk. The clear central area was

*In the group of nine eggs I deduce this from Mr. Jeffrey's observations communicated by letter.

therefore much increased in size, but not so limpidly transparent as the (empty?) space surrounding the annulus externally. Immediately within the annulus, on the morning of the 7th, I observed the clear area all round to the peduncle or umbilicus on both sides, to have its outer edge crenated. In the more advanced eggs these crenations soon developed into a row of (ventral) segments, twelve in number, marked off from the rest of the central area by a curved concentric line. These segments had an optical area from four to six times as large as the reticulations of the chorion (say about inch in diameter). Soon three of them, at the south extremity of the series, appeared larger than the rest and furnished with processes sloping northward; whilst still further to the south, but less distinctly, two or three other segments could be made out having also processes sloping in the opposite direction. Into this space, corresponding to the region between the head and the body, faint cloudlets of yolk-granules appeared as it were passing in from the annulus towards the central area. Owing probably to the sloping in opposite directions of the thoracic and cephalic processes, there was less pressure here of the amnion against the serous membrane and freer passage for the yolk. This appearance at this region was visible persistently in all the eggs (Fig. 34). The thoracic processes or legs were about half the width of the segments from the posterior half of which they took their origin, but considerably longer, passing out of sight in the annulus (Fig. 34). About this time I noticed also other appearances as of curved concentric lines marking out a tube (mesenteron?) but which would require further observation for their certain interpretation. Upon the same day (Aug. 7) in the afternoon, I noticed eye spots (in all but N.W.) a group of about six arranged in a circular form (Fig. 35). There were now visible at least four cephalic segments, and it was at the base of the third (reckoning from behind forwards) that the group of eye-spots was situated. Later on I saw a fifth larger terminal mass forming the extreme anterior extremity of the embryo. The thoracic legs appeared to be jointed. At 2.50 P.M. I noticed that the twelfth (ninth abdominal) segment was somewhat longer than the others and projected inwards towards the centre, beyond their level. At 6.55 P.M. this inward projection had disappeared, and a very important change in the terminal segment had been initiated. It had become ventrally incurved upon itself (Fig. 36). This segment was elongated, and narrowed at the apex. It advanced steadily forwards along the ventral aspect of the embryo, followed by the others, and growing larger at the same time; but the head remained always in its original place. On the morning of the 8th the last three segments (10-12) were round the corner (Fig. 37); at 9.10 A.M. the tail was in contact with the metathoracic legs; at 12.45 P.M. it had gone beyond the two posterior pairs of legs, which, with the forelegs, were now directed

forwards in place of backwards as at first. The middle of the bend of the abdomen (at the incurvature) was between the third and fourth abdominal segments; at I P.M. the tail had come quite up with the forefeet; at 2.40 P.M. it reached fully up to the head and fully filled the larger northern bay. At this time the stump of the umbilicus on the one hand, now reduced to a mere conical point, and a similar projection from the opposite side of the annulus, where cloudlets of yolk-granules were passing in towards the neck region (see Fig. 34), indicated a division of the whole interior space into two unequal bays, of which the smaller (southern) contained the head, and the other the rest of the body. The second abdominal segment might be said to form the keystone of the arch where the abdomen was bent on itself, but the head, though free in its bay, remained mostly in close contact with the umbilical stalk. On the morning of the 9th the tail reached quite to the level of the head and beyond the eye-spots. The embryo lay in a loop with the legs inside, a position which it had reached by the growth and enlargement of the tail without any change of place in the head and anterior segments.

I have dwelt at some length on this incurvature and growth of the tail by which the lepidopterous embryo attains the loop-form in the shell, because Kowalevski has stated (Mém. de l'Acad. Imp. des Sciences de St. Pétersbourg, vii. série, tome xvi. No. 12, p. 56), that it does so by the whole embryo turning round in the shell after its tail. "Dem Hinterende folgend, dreht sich der ganze Embryo so, dass er jetzt der ihn noch bedeckenden serösen Hülle den Rücken zuwendet, und die Extremitäten erscheinen nach innen gerichtet." Perhaps the subject may be made clearer by a brief consideration of the different kinds of motion which may be observed in eggs. These may be classed under four heads. Ist. Movements due to gravitation. The ventral or developing side of the yolk in the egg of Gastrophysa raphani, e.g., turns always towards the upper surface, though this change takes place so slowly that it may occupy several days in completion. 2nd. Movements of growth; strikingly illustrated in the egg of Calopteryx, in which the embryo becomes inverted in the shell (v. Balfour, Comp. Embryol. i. 334). 3rd. Embryonic movements; by which limbs or parts show movements without any change of place in the whole; and, lastly, larval movements; when the perfectly formed embryo changes its position in the shell, or acts in any other way as if it were independent of it. The loop form of the lepidopterous embryo, Kowalevski supposes to be due to the latter class of movements, whilst in reality it is only a movement of growth. When, in its final stages, as stated by Kowalevski and as observed in these eggs by Mr. Jeffrey, the embryo of Botys devours the remainder of the yolk and cuts its way out of the shell, these actions may be fairly described as larval movements.

On the afternoon of the &h August, I had a pretty clear view of the head mobscured by the annulus. The four processes were approximated like the fingers of the hand, whilst the fifth or terminal lobe lay away from ther like a thumb (Fig. 38). Projecting into the bay between the thumb and the fingers, and crossing the latter obliquely, there was faintly visible another process which I was unable to follow further. On the other side of the umbilical stalk, now worn down to a conical stump, a dense crescentic streak of yoke had been as it were detached and carried away by the growth of the tail till it came to occupy the northern curve of the bay. This included yolk, according to Kowalevski, marks the extent of the mesenteron, in this case about a fourth of the whole alimentary canal. The incurved portion of the tail was free from the annulus, and its growth seemed to be caused by the formation of the dorsal half of the body carrying along the ventral segments with it. In the tail this dorsal portion increased much in size beyond the ventral, both in length and thickness. Up till its full growth the dorsal section of the abdomen had been without segmentation, but, in the afternoon of the 8th August, I observed it to be crenate externally; and that the indentations corresponded to the lines of division of the ventral segments. That portion of the dorsum extending beyond the venter and beyond the last crenation opposite the last ventral segment, was much larger than any other division, and showed itself a crenate division into three, of which the last, much narrower than the others, I supposed might turn out to be the anal proleg, whilst the penult crenation would represent the anal flap. I thought also I could trace the posterior section of the alimentary canal terminating in the space between the anal flap (?) and the proleg (?) In the evening of the same day I was able to distinguish four prolegs on the anterior abdominal segments, near the point of flexure, but could not see whether one or two segments intervened between them and the thoracic legs. On the 8th also, in the evening, the remote group of eye-spots became dimly visible through the transparent head. At first, directly opposite one another, the two groups of eye-spots diverged more and more till they came to be situated at the lateral borders of the head, and between, and rather in advance of them, a cupid's-bow-shaped line (Fig. 39) appeared to indicate the anterior border of the clypeus. The direction of this torsion of the head was always such that, whether the embryo was lying on the right or left side, its effect was to bring the dorsal aspect of the head next the free unattached side of the egg, and the under surface next the glass. Unfortunately my observations were brought to an untimely close by an accident on the morning of the 9th, so that the last stages, as well as the earlier, both of which should have much of interest to offer, escaped me. I believe, however, that what I have

been fortunate enough, thanks to Mr. Jeffrey's kindness, to see, is not without some interest and importance from its bearing on two points: the orientation of the embryo in the shell, and the incurvature of the tail. If there were any certainty of obtaining similar eggs another season, I would have reserved some of the other points for further observation as it is, I cannot refrain from mentioning the hypothesis that, as plants are bent to or from the light by a preponderance of growth on the opposite side, so, here, the proximate cause of the ventral curvature of the tail end is the later, but then quicker and predominant growth of the dorsal section of the embryo.

Milford, Letterkenny.

SOME NEW

H

DIATOMACEOUS

FROM THE

DUBRAVICA.

FORMS

"SAUGSCHIEFER" OF

By F. KITTON, Hon. F.R.M.S.

ERR GRUNOW in his "

'Beiträge zur Kentniss der fossilen Diatomaceen OesterreichUngarn" ("Beiträge zur paläontol. Oester-Ungarns und d. Orients von Majsisovics u. Naumayr," Bd. ii. 1882), describes the following diatomaceous deposits found in Hungary: (1) "Saugschiefer " (absorbent slate) from Dubrávica; (2) Polierschiefer Tallya; (3) the argillaceous tufa from Holaikluk; (4) diatom deposit from Kis-ker; (5) Kieselguhre from Eger and Franzenbad in Bohemia. The last named deposits are generally well known to Diatomists (particularly that found in Franzenbad). Ehrenberg described many of the forms in the "Monatsberichte" of the Royal Academy of Berlin, 1840, which are afterwards figured in his "Microgeologie." His figures of Campylodiscus clypeus have been frequently copied.

The Hungarian deposits were almost, if not entirely, unknown until the recently published investigations of Herr Grunow, which, unfortunately, are not readily accessible to Diatomists, excepting. by the purchase of the volume of the work (at a cost of 40 marks) in which they appeared. Having through the kindness of a correspondent been enabled to examine one of the most interesting of them, viz. that from Dubrávica, I have identified most of the species named in the list which accompanied the sample. The deposit is somewhat delusive; from its general appearance we should suppose it would be easily cleaned; but this is not the case; when boiled in acids the material split up into thin laminæ of sufficient tenuity to allow of mounting without further manipulation; to separate the diatoms a careful boil in dehydrated soda is necessary to dissolve the silicic acid which cements the diatoms together. This cementation