Chapter XVII: Tracheata (3)

(410) Ganin. "Beiträge zur Erkenntniss d. Entwicklungsgeschichte d. Insecten." _Zeit. f. wiss. Zool._ Bd. XIX. 1869.

(411) V. Graber. _Die Insecten._ München, 1877.

(412) V. Graber. "Vorlauf. Ergeb. üb. vergl. Embryologie d. Insecten." _Archiv f. mikr. Anat._ Vol. XV. 1878.

(413) O. v. Grimm. "Ungeschlechtliche Fortpflanzung einer Chironomus-Art. u. deren Entwicklung aus dem unbefruchteten Ei." _Mém. Acad. Pétersbourg._ 1870.

(414) B. Hatschek. "Beiträge zur Entwicklung d. Lepidopteren." _Jenaische Zeitschrift_, Bd. XI.

(415) A. Kölliker. "Observationes de primâ insectorum genese, etc." _Ann. Sc. Nat._ Vol. XX. 1843.

(416) A. Kowalevsky. "Embryologische Studien an Würmern u. Arthropoden." _Mém. Ac. imp. Pétersbourg_, Ser. VII. Vol. XVI. 1871.

(417) C. Kraepelin. "Untersuchungen üb. d. Bau, Mechanismus u. d. Entwick, des Stachels d. bienartigen Thiere." _Zeit. f. wiss. Zool._ Vol. XXIII. 1873.

(418) C. Kupffer. "Faltenblatt an d. Embryonen d. Gattung Chironomus." _Arch. f. mikr. Anat._ Vol. II. 1866.

(419) R. Leuckart. _Zur Kenntniss d. Generationswechsels u. d. Parthenogenese b. d. Insecten._ Frankfurt, 1858.

(420) Lubbock. _Origin and Metamorphosis of Insects._ 1874.

(421) Lubbock. _Monograph on Collembola and Thysanura._ Ray Society, 1873.

(422) Melnikow. "Beiträge z. Embryonalentwicklung d. Insecten." _Archiv f. Naturgeschichte_, Bd. XXXV. 1869.

(423) E. Metschnikoff. "Embryologische Studien an Insecten." _Zeit. f. wiss. Zool._ Bd. XVI. 1866.

(424) P. Meyer. "Ontogenie und Phylogenie d. Insecten." _Jenaische Zeitschrift_, Vol. X. 1876.

(425) Fritz Müller. "Beiträge z. Kenntniss d. Termiten." _Jenaische Zeitschrift_, Vol. IX. 1875.

(426) A. S. Packard. "Embryological Studies on Diplex, Perithemis, and the Thysanurous genus Isotoma." _Mem. Peabody Acad. Science_, 1. 2. 1871.

(427) Suckow. "Geschlechtsorgane d. Insecten." Heusinger's _Zeitschrift f. organ. Physik_, Bd. II. 1828.

(428) Tichomiroff. "Ueber die Entwicklungsgeschichte des Seidenwurms." _Zoologischer Anzeiger_, II. Jahr. No. 20 (Preliminary Notice).

(429) Aug. Weismann. "Zur Embryologie d. Insecten." _Archiv f. Anat. und Phys._ 1864.

(430) Aug. Weismann. "Entwicklung d. Dipteren." _Zeit. f. wiss. Zool._ Vols. XIII. and XIV. Leipzig, 1863-4.

(431) Aug. Weismann. "Die Metamorphose d. Corethra plumicornis." _Zeit. f. wiss. Zool._ Vol. XVI. 1866.

(432) N. Wagner. "Beitrag z. Lehre d. Fortpflanzung d. Insectenlarven." _Zeit. f. wiss. Zool._ Vol. XIII. 1860.

(433) Zaddach. _Untersuchungen üb. d. Bau u. d. Entwicklung d. Gliederthiere._ Berlin, 1854.

ARACHNIDA[179].

[179] The classification of the Arachnida adopted in the present
work is shewn below:
I. Arthrogastra. {Scorpionidæ.
{Pedipalpi.
{Pseudoscorpionidæ.
{Solifugæ.
{Phalangidæ.
II. Araneina. {Tetrapneumones.
{Dipneumones.
III. Acarina.

The development of several divisions of this interesting group has been worked out; and it will be convenient to deal in the first instance with the special history of each of these divisions, and then to treat in a separate section the development of the organs for the whole group.

[FIG. 193. OVUM OF SCORPION WITH THE ALREADY FORMED BLASTODERM
SHEWING THE PARTIAL SEGMENTATION. (After Metschnikoff.)

_bl_. blastoderm.]

Scorpionidæ. The embryonic development always takes place within the female Scorpion. In Buthus it takes place within follicle-like protuberances of the wall of the ovary. In Scorpio also development commences while the egg is still in the follicle, but when the trunk becomes segmented the embryo passes into the ovarian tube. The chief authority for the development of the Scorpionidæ is Metschnikoff (No. 434).

At the pole of the ovum facing the ovarian tube there is formed a germinal disc which undergoes a partial segmentation (fig. 193 _bl_). A somewhat saucer-shaped one-layered blastoderm is then formed, which soon becomes thickened in the centre and then divided into two layers. The outer of these is the epiblast, the inner the mesoblast. Beneath the mesoblast there subsequently appear granular cells, which form the commencement of the hypoblast[180].

[180] The origin of the hypoblast cells, if such these cells are,
is obscure. Metschnikoff doubtfully derives them from the
blastoderm cells; from my investigations on Spiders it appears to
me more probable that they originate in the yolk.

During the formation of the blastoderm a cellular envelope is formed round the embryo. Its origin is doubtful, though it is regarded by Metschnikoff as probably derived from the blastoderm and homologous with the amnion of Insects. It becomes double in the later stages (fig. 195).

[FIG. 194. THREE SURFACE VIEWS OF THE VENTRAL PLATE OF A DEVELOPING
SCORPION. (After Metschnikoff.)

A. Before segmentation.
B. After five segments have become formed.
C. After the appendages have begun to be formed.]

During the differentiation of the three embryonic layers the germinal disc becomes somewhat pyriform, the pointed end being the posterior. At this extremity there is a special thickening which is perhaps equivalent to the primitive cumulus of Spiders. The germinal disc continues gradually to spread over the yolk, but the original pyriform area is thicker than the remainder, and is marked off anteriorly and posteriorly by a shallow furrow. It constitutes a structure corresponding with the ventral plate of other Tracheata. It soon becomes grooved by a shallow longitudinal furrow (fig. 194 A) which subsequently becomes less distinct. It is then divided by two transverse lines into three parts.[181]

[181] The exact fate of the three original segments is left
somewhat obscure by Metschnikoff. He believes however that the
anterior segment forms the procephalic lobes, the posterior
probably the telson and five adjoining caudal segments, and the
middle one the remainder of the body. This view does not appear
to me quite satisfactory, since on the analogy of Spiders and
other Arthropoda the fresh somites ought to be added by a
continuous segmentation of the posterior lobe.

In succeeding stages the anterior of the three parts is clearly marked out as the procephalic lobe, and soon becomes somewhat broader. Fresh segments are added from before backwards, and the whole ventral plate increases rapidly in length (fig. 194 B).

When ten segments have become formed, appendages appear as paired outgrowths of the nine posterior segments (fig. 194 C). The second segment bears the pedipalpi, the four succeeding segments the four ambulatory appendages, and the four hindermost segments smaller provisional appendages which subsequently disappear, with the possible exception of the second. The foremost segment, immediately behind the procephalic lobes, is very small, and still without a rudiment of the cheliceræ, which are subsequently formed on it. It would appear from Metschnikoff's figures to be developed later than the other post-oral segments present at this stage. The still unsegmented tail has become very prominent and makes an angle of 180° with the remainder of the body, over the ventral surface of which it is flexed.

[FIG. 195. A FAIRLY ADVANCED EMBRYO OF THE SCORPION ENVELOPED IN ITS
MEMBRANES. (After Metschnikoff.)

_ch._ cheliceræ; _pd._ pedipalpi; _p1-p4._ ambulatory appendages;
_ab._ post-abdomen (tail).]

By the time that twelve segments are definitely formed, the procephalic region is distinctly bilobed, and in the median groove extending along it the stomodæum has become formed (fig. 196 A). The cheliceræ (ch) appear as small rudiments on the first post-oral segment, and the nerve cords are distinctly differentiated and ganglionated. In the embryonic state there is one ganglion for each segment. The ganglion in the first segment (that bearing the cheliceræ) is very small, but is undoubtedly post-oral.

At this stage, by a growth in which all the three germinal layers have a share, the yolk is completely closed in by the blastoderm. It is a remarkable fact with only few parallels, and those amongst the Arthropoda, that the blastopore, or point where the embryonic membranes meet in closing in the yolk, is situated on the dorsal surface of the embryo.

The general relations of the embryo at about this stage are shewn in fig. 195, where the embryo enclosed in its double cellular membrane is seen in a side view. This embryo is about the same age as that seen from the ventral surface in fig. 196 A.

The general nature of the further changes may easily be gathered from an inspection of fig. 196 B and C, but a few points may be noted.

An upper lip or labrum is formed as an unpaired organ in the line between the procephalic lobes. The pedipalpi become chelate before becoming jointed, and the cheliceræ also early acquire their characteristic form. Rudimentary appendages appear on the six segments behind the ambulatory legs, five of which are distinctly shewn in fig. 195; they persist only on the second segment, where they appear to form the comb-like organs or pectines. The last abdominal segment, _i.e._ that next the tail, is without provisional appendages. The embryonic tail is divided into six segments including the telson (fig. 196 C, _ab_). The lungs (_st_) are formed by paired invaginations, the walls of which subsequently become plicated, on the four last segments which bear rudimentary limbs, and simultaneously with the disappearance of the rudimentary limbs.

[FIG. 196. THREE STAGES IN THE DEVELOPMENT OF THE SCORPION. THE
EMBRYOS ARE REPRESENTED AS IF SEEN EXTENDED ON A PLANE. (After
Metschnikoff.)

_ch._ cheliceræ; _pd._ pedipalpi; _p1-p4._ ambulatory appendages;
_pe._ pecten; _st._ stigmata; _ab._ post abdomen (tail).]

Pseudoscorpionidæ. The development of Chelifer has been investigated by Metschnikoff (436), and although (except that it is provided with tracheæ instead of pulmonary sacks) it might be supposed to be closely related to Scorpio, yet in its development is strikingly different.

The eggs after being laid are carried by the female attached to the first segment of the abdomen. The segmentation (_vide_ p. 93) is intermediate between the types of complete and superficial segmentation. The ovum, mainly formed of food-yolk, divides into two, four, and eight equal segments (fig. 197 A). There then appear one or more clear segments on the surface of these, and finally a complete layer of cells is formed round the central yolk spheres (fig. 197 B), which latter subsequently agglomerate into a central mass. The superficial cells form what may be called a blastoderm, which soon becomes divided into two layers (fig. 197 C). There now appears a single pair of appendages (the pedipalpi) (fig. 198 A, _pd_), while at the same time the front end of the embryo grows out into a remarkable proboscis-like prominence--a temporary upper lip (concealed in the figure behind the pedipalpus), and the abdomen (_ab_) becomes bent forwards towards the ventral surface. In this very rudimentary condition, after undergoing an ecdysis, the larva is hatched, although it still remains attached to its parent. After hatching it grows rapidly, and becomes filled with a peculiar transparent material. The first pair of ambulatory appendages is formed behind the pedipalpi and then the three succeeding pairs, while at the same time the cheliceræ appear as small rudiments in front. External signs of segmentation have not yet appeared, but about this period the nervous system is formed. The supra-oesophageal ganglia are especially distinct, and provided with a central cavity, probably formed by an invagination, as in other Arachnida. In the succeeding stages (fig. 198 B) four provisional pairs of appendages (shewn as small knobs at _ab_) appear behind the ambulatory feet. The abdomen is bent forwards so as to reach almost to the pedipalpi. In the later stages (fig. 198 C) the adult form is gradually attained. The enormous upper lip persists for some time, but subsequently atrophies and is replaced by a normal labrum. The appendages behind the ambulatory feet atrophy, and the tail is gradually bent back into its final position. The segmentation and the gradual growth of the limbs do not call for special description, and the formation of the organs, so far as is known, agrees with other types.

[FIG. 197. SEGMENTATION AND FORMATION OF THE BLASTODERM IN CHELIFER.
(After Metschnikoff.)

In A the ovum is divided into a number of separate segments. In B a
number of small cells have appeared (_bl_) which form a blastoderm
enveloping the large yolk spheres. In C the blastoderm has become
divided into two layers.]

The segmentation of Chthonius is apparently similar to that of Chelifer (Stecker, No. 437).

Phalangidæ. Our knowledge of the development of Phalangium is unfortunately confined to the later stages (Balbiani, No. 438). These stages do not appear however to differ very greatly from those of true Spiders.

Araneina. The eggs of true Spiders are either deposited in nests made specially for them, or are carried about by the females. Species belonging to a considerable number of genera, viz. Pholcus, Epeira, Lycosa, Clubione, Tegenaria and Agelena have been studied by Claparède (No. 442), Balbiani (No. 439), Barrois (No. 441) and myself (No. 440), and the close similarity between their embryos leaves but little doubt that there are no great variations in development within the group.

[FIG. 198. THREE STAGES IN THE DEVELOPMENT OF CHELIFER. (After
Metschnikoff.)

_pd._ pedipalpi; _ab._ abdomen; _an.i._ anal invagination; _ch._
cheliceræ.]

The ovum is enclosed in a delicate vitelline membrane, enveloped in its turn by a chorion secreted by the walls of the oviduct. The chorion is covered by numerous rounded prominences, and occasionally exhibits a pattern corresponding with the areas of the cells which formed it. The segmentation has already been fully described, pp. 118 and 119. At its close there is present an enveloping blastoderm formed of a single layer of large flattened cells. The yolk within is formed of a number of large polygonal segments; each of which is composed of large yolk-spherules, and contains a nucleus surrounded by a layer of protoplasm, which is prolonged into stellate processes holding together the yolk-spherules. The nucleus, surrounded by the major part of the protoplasm of each yolk cell, appears, as a rule, to be situated not at the centre, but on one side of its yolk segment.

The further description of the development of Spiders applies more especially to Agelena labyrinthica, the species which formed the subject of my own investigations.

The first differentiation of the blastoderm consists in the cells of nearly the whole of one hemisphere becoming somewhat more columnar than those of the other hemisphere, and in the cells of a small area near one end of the thickened hemisphere becoming distinctly more columnar than elsewhere, and two layers thick. This area forms a protuberance on the surface of the ovum, originally discovered by Claparède, and called by him the primitive cumulus. In the next stage the cells of the thickened hemisphere of the blastoderm become still more columnar; and a second area, at first connected by a whitish streak with the cumulus, makes its appearance. In the second area the blastoderm is also more than one cell deep (fig. 199). It will be noticed that the blastoderm, though more than one cell thick over a large part of the ventral surface, is not divided into distinct layers. The second area appears as a white patch and soon becomes more distinct, while the streak continued to it from the cumulus is no longer visible. It is shewn in surface view in fig. 200 A. Though my observations on this stage are not quite satisfactory, yet it appears to me probable that there is a longitudinal thickened ridge of the blastoderm extending from the primitive cumulus to the large white area. The section represented in fig. 199, which I believe to be oblique, passes through this ridge at its most projecting part.

[FIG. 199. SECTION THROUGH THE EMBRYO OF AGELENA LABYRINTHICA.

The section is from an embryo of the same age as fig. 200 A, and is
represented with the ventral plate upwards. In the ventral plate
is seen a keel-like thickening, which gives rise to the main mass
of the mesoblast.
_yk._ yolk divided into large polygonal cells, in several of which
nuclei are shewn.]

The nuclei of the yolk cells during the above stages multiply rapidly, and cells are formed in the yolk which join the blastoderm; there can however be no doubt that the main increase in the cells of the blastoderm has been due to the division of the original blastoderm cells.

In the next stage I have been able to observe there is, in the place of the previous thickened half of the blastoderm, a well-developed ventral plate with a procephalic lobe in front, a caudal lobe behind, and an intermediate region marked by about three transverse grooves, indicating a division into segments. This plate is throughout two or more rows of cells thick, and the cells which form it are divided _into two distinct layers_--a columnar superficial layer of epiblast cells, and a deeper layer of mesoblast cells (fig. 203 A). In the latter layer there are several very large cells which are in the act of passing from the yolk into the blastoderm. The identification of the structures visible in the previous stage with those visible in the present stage is to a great extent a matter of guess-work, but it appears to me probable that the primitive cumulus is still present as a slight prominence visible in surface views on the caudal lobe, and that the other thickened patch persists as the procephalic lobe. However this may be, the significance of the primitive cumulus appears to be that it is the part of the blastoderm where two rows of cells become first established[182].

[182] Various views have been put forward by Claparède and
Balbiani about the position and significance of the primitive
cumulus. For a discussion of which _vide_ self, No. 440.

The whole region of the blastoderm other than the ventral plate is formed of a single row of flattened epiblast cells. The yolk retains its original constitution.

By this stage the epiblast and mesoblast are distinctly differentiated, and the homologue of the hypoblast is to be sought for in the yolk cells. The yolk cells are not however entirely hypoblastic, since they continue for the greater part of the development to give rise to fresh cells which join the mesoblast.

The Spider's blastoderm now resembles that of an Insect (except for the amnion) after the establishment of the mesoblast, and the mode of origin of the mesoblast in both groups is very similar, in that the longitudinal ridge-like thickening of the mesoblast shewn in fig. 199 is probably the homologue of the mesoblastic groove of the Insects' blastoderm.

The ventral plate continues to grow rapidly, and at a somewhat later stage (fig. 200 B) there are six segments interposed between the procephalic and caudal lobes. The two anterior of these (_ch_ and _pd_), especially the foremost, are less distinct than the remainder; and it is probable that both of them, and in any case the anterior one, are formed later than the three segments following. These two segments are the segments of the cheliceræ and pedipalpi. The four segments following belong to the four pairs of ambulatory legs. The segments form raised transverse bands separated by transverse grooves. There is at this stage a faintly marked groove extending along the median line of the ventral plate. This groove is mainly caused by the originally single mesoblastic plate having become divided throughout the whole region of the ventral plate, except possibly the procephalic lobes, into two bands, one on each side of the middle line (fig. 203 B).

[FIG. 200. FOUR STAGES IN THE DEVELOPMENT OF AGELENA LABYRINTHICA.

A. Stage when the ventral plate is very imperfectly differentiated.
_pr.c._ primitive cumulus.
B. Ovum viewed from the side when the ventral plate has become
divided into six segments. _ch._ segment of cheliceræ imperfectly
separated from procephalic lobe; _pd._ segment of pedipalpi.
C. Ventral plate ideally unrolled after the full number of segments
and appendages are established. _st._ stomodæum between the two
præ-oral lobes. Behind the six pairs of permanent appendages are
seen four pairs of provisional appendages.
D and E. Two views of an embryo at the same stage. D ideally
unrolled, E seen from the side. _st._ stomodæum; _ch._ cheliceræ;
on their inner side is seen the ganglion belonging to them. _pd._
pedipalpi; _pr.p._ provisional appendages.]

The segments continue to increase in number by the continuous addition of fresh segments between the one last formed and the caudal lobe. By the stage with nine segments the first rudiments of the limbs make their appearance. The first rudiments to appear are those of the pedipalpi and four ambulatory limbs: the cheliceræ, like the segment to which they belong, lag behind in development. The limbs appear as small protuberances at the borders of their segments. By the stage when they are formed the procephalic region has become bilobed, and the two lobes of which it is composed are separated by a shallow groove.

By a continuous elongation the ventral plate comes to form a nearly complete equatorial ring round the ovum, the procephalic and caudal lobes being only separated by a very narrow space, the undeveloped dorsal region of the embryo. This is shewn in longitudinal section in fig. 204. In this condition the embryo may be spoken of as having a _dorsal flexure_. By the time that this stage is reached (fig. 200 C) the full number of segments and appendages has become established. There are in all sixteen segments (including the caudal lobe). The first six of these bear the permanent appendages of the adult; the next four are provided with provisional appendages; while the last six are without appendages. The further features of this stage which deserve notice are (1) the appearance of a shallow depression (_st_)--the rudiment of the stomodæum--between the hinder part of the two procephalic lobes; (2) the appearance of raised areas on the inner side of the six anterior appendage-bearing segments. These are the rudiments of the ventral ganglia. It deserves to be especially noted that the segment of the cheliceræ, like the succeeding segments, is provided with ganglia; and that the ganglia of the cheliceræ are quite distinct from the supra-oesophageal ganglia derived from the procephalic lobes. (3) The pointed form of the caudal lobe. In Pholcus (Claparède, No. 442) the caudal lobe forms a projecting structure which, like the caudal lobe of the Scorpion, bends forward so as to face the ventral surface of the part of the body immediately in front. In most Spiders such a projecting caudal lobe is not found. While the embryo still retains its dorsal flexure considerable changes are effected in its general constitution. The appendages (fig. 200 D and E) become imperfectly jointed, and grow inwards so as to approach each other in the middle line. Even in the stage before this, the ventral integument between the rudiments of the ganglia had become very much thinner, and had in this way divided the ventral plate into two halves. At the present stage the two halves of the ventral plate are still further separated, and there is a wide space on the ventral side only covered by a delicate layer of epiblast. This is shewn in surface view (fig. 200 D) and in section in fig. 203 C.

The stomodæum (_st_) is much more conspicuous, and is bounded in front by a prominent upper lip, and by a less marked lip behind. The upper lip becomes less conspicuous in later stages, and is perhaps to be compared with the provisional upper lip of Chelifer. Each procephalic lobe is now marked by a deep semicircular groove.

The next period in the development is characterised by the gradual change in the flexure of the embryo from a dorsal to a ventral one; accompanied by the division of the body into an abdomen and cephalo-thorax, and the gradual assumption of the adult characters.

The change in the flexure of the embryo is caused by the elongation of the dorsal region, which has hitherto been hardly developed. Such an elongation increases the space on the dorsal surface between the procephalic and caudal regions, and therefore necessarily separates the caudal and procephalic lobes; but, since the ventral plate does not become shortened in the process, and the embryo cannot straighten itself in the egg-shell, it necessarily becomes ventrally flexed.

If there were but little food-yolk this flexure would naturally cause the whole embryo to be bent in so as to have the ventral surface concave. But instead of this the flexure is at first confined to the two bands which form the ventral plate. These bands, as shewn in fig. 201 A, acquire a true ventral flexure, but the yolk forms a projection--a kind of yolk-sack as Barrois (No. 441) calls it--distending the thin integument between the two ventral bands. This yolk-sack is shewn in surface view in fig. 201 A and in section in fig. 206. At a later period, when the yolk has become largely absorbed, the true nature of the ventral flexure becomes quite obvious, since the abdomen of the young Spider, while still in the egg, is found to be bent over so as to press against the ventral surface of the thorax (fig. 201 B).

[FIG. 201. TWO LATE STAGES IN THE DEVELOPMENT OF AGELENA
LABYRINTHICA.

A. Embryo from the side at the stage when there is a large ventral
protuberance of yolk. The angle between the line of insertion of
the permanent and provisional appendages shews the extent of the
ventral flexure.
B. Embryo nearly ready to be hatched. The abdomen which has not
quite acquired its permanent form is seen to be pressed against
the ventral side of the thorax.
_pr.l._ procephalic lobe; _pd._ pedipalpi; _ch._ cheliceræ; _c.l._
caudal lobe; _pr.p._ provisional appendages.]

The general character of the changes which take place during this period in the development is shewn in fig. 201 A and B representing two stages in it. In the first of these stages there is no constriction between the future thorax and abdomen. The four pairs of provisional appendages exhibit no signs of atrophy; and the extent of the ventral flexure is shewn by the angle formed between the line of their insertion and that of the appendages in front. The yolk has enormously distended the integument between the two halves of the ventral plate, as is illustrated by the fact that, at a somewhat earlier stage than that figured, the limbs cross each other in the median ventral line, while at this stage they do not nearly meet. The limbs have acquired their full complement of joints, and the pedipalpi bear a cutting blade on their basal joint.

The dorsal surface between the prominent caudal lobe and the procephalic lobes forms more than a semicircle. The terga are fully established, and the boundaries between them, especially in the abdomen, are indicated by transverse markings. A large lower lip now bounds the stomodæum, and the upper lip has somewhat atrophied. In the later stage (fig. 201 B) the greater part of the yolk has passed into the abdomen, which is now to some extent constricted off from the cephalo-thorax. The appendages of the four anterior abdominal somites have disappeared, and the caudal lobe has become very small. In front of it are placed two pairs of spinning mammillæ. A delicate cuticle has become established, which is very soon moulted.

Acarina. The development of the Acarina, which has been mainly investigated by Claparède (No. 446), is chiefly remarkable from the frequent occurrence of several larval forms following each other after successive ecdyses. The segmentation (_vide_ p. 116) ends in the formation of a blastoderm of a single layer of cells enclosing a central yolk mass.

A ventral plate is soon formed as a thickening of the blastoderm, in which an indistinct segmentation becomes early observable. In Myobia, which is parasitic on the common mouse, the ventral plate becomes divided by five constrictions into six segments (fig. 202 A), from the five anterior of which paired appendages very soon grow out (fig. 202 B). The appendages are the cheliceræ (_ch_) and pedipalpi (_pd_) and the first three pairs of limbs (_p1-p3_). On the dorsal side of the cheliceræ a thickened prominence of the ventral plate appears to correspond to the procephalic lobes of other Arachnida. The part of the body behind the five primitive appendage-bearing segments appears to become divided into at least two segments. In other mites the same appendages are formed as in Myobia, but the preceding segmentation of the ventral plate is not always very obvious.

In Myobia two moultings take place while the embryo is still within the primitive egg-shell. The first of these is accompanied by the _apparently total disappearance of the three pediform appendages_, and the complete coalescence of the two gnathiform appendages into a proboscis (fig. 202 C). The feet next grow out again, and a second ecdysis then takes place. The embryo becomes thus inclosed within three successive membranes, viz. the original egg-shell and two cuticular membranes (fig. 202 D). After the second ecdysis the appendages assume their final form, and the embryo leaves the egg as an hexapodous larva. The fourth pair of appendages is acquired by a post-embryonic metamorphosis. From the proboscis are formed the rudimentary palpi of the second pair of appendages, and two elongated needles representing the cheliceræ.

[FIG. 202. FOUR SUCCESSIVE STAGES IN THE DEVELOPMENT OF MYOBIA
MUSCULI. (After Claparède.)

_s1-s4._ post-oral segments; _ch._ cheliceræ; _pd._ pedipalpi; _pr_
proboscis formed by the coalescence of the cheliceræ and pedipalpi;
_p1, p2_, etc. ambulatory appendages.]

In the cheese mite (Tyroglyphus) the embryo has two ecdyses which are not accompanied by the peculiar changes observable in Myobia: the cheliceræ and pedipalpi fuse however to form the proboscis. The first larval form is hexapodous, and the last pair of appendages is formed at a subsequent ecdysis.

In Atax Bonzi, a form parasitic on Unio, the development and metamorphosis are even more complicated than in Myobia. The first ecdysis occurs before the formation of the limbs, and shortly after the ventral plate has become divided into segments. Within the cuticular membrane resulting from the first ecdysis the anterior five pairs of limbs spring out in the usual fashion. They undergo considerable differentiation; the cheliceræ and pedipalpi approaching each other at the anterior extremity of the body, and the three ambulatory legs becoming segmented and clawed. An oesophagus, a stomach, and an oesophageal nerve-ring are also formed. When the larva has attained this stage the original egg-shell is split into two valves and eventually cast off, but the embryo remains enclosed within the cuticular membrane shed at the first ecdysis. This cuticular membrane is spoken of by Claparède as the deutovum. In the deutovum the embryo undergoes further changes; the cheliceræ and pedipalpi coalesce and form the proboscis; a spacious body cavity with blood corpuscles appears; and the alimentary canal enclosing the yolk is formed.

The larva now begins to move, the cuticular membrane enclosing it is ruptured, and the larva becomes free. It does not long remain active, but soon bores its way into the gills of its host, undergoes a fresh moult, and becomes quiescent. The cuticular membrane of the moult just effected swells up by the absorption of water and becomes spherical. Peculiar changes take place in the tissues, and the limbs become, as in Myobia, nearly absorbed, remaining however as small knobs. The larva swims about as a spherical body within its shell. The feet next grow out afresh, and the posterior pair is added. From the proboscis the palpi (of the pedipalpi) grow out below. The larva again becomes free, and amongst other changes the cheliceræ grow out from the proboscis. A further ecdysis, with a period of quiescence, intervenes between this second larval form and the adult state.

The changes in the appendages which appear common to the Mites generally are (1) the late development of the fourth pair of appendages, which results in the constant occurrence of an hexapodous larva; and (2) the early fusion of the cheliceræ and pedipalpi to form a proboscis in which no trace of the original appendages can be discerned. In most instances palpi and stilets of variable form are subsequently developed in connexion with the proboscis, and, as indicated in the above descriptions, are assumed to correspond with the two original embryonic appendages.

_The history of the germinal layers._

It is a somewhat remarkable fact that each of the groups of the Arachnida so far studied has a different form of segmentation. The types of Chelifer and the Spiders are simple modifications of the centrolecithal type, while that of Scorpio, though apparently meroblastic, is probably to be regarded in the same light (_vide_ p. 120 and p. 434). The early development begins in the Scorpion and Spiders with the formation of a ventral plate, and there can be but little doubt that Chelifer is provided with an homologous structure, though very probably modified, owing to the small amount of food-yolk and early period of hatching.

The history of the layers and their conversion into the organs has been studied in the case of the Scorpion (Metschnikoff, No. 434), and of the Spiders; and a close agreement has been found to obtain between them.

It will be convenient to take the latter group as type, and simply to call attention to any points in which the two groups differ.

The epiblast. The epiblast, besides giving rise to the skin (hypodermis and cuticle), also supplies the elements for the nervous system and organs of sense, and for the respiratory sacks, the stomodæum and proctodæum.

At the period when the mesoblast is definitely established, the epiblast is formed of a single layer of columnar cells in the region of the ventral plate, and of a layer of flat cells over other parts of the yolk.

When about six segments are present the first changes take place. The epiblast of the ventral plate then becomes somewhat thinner in the median line than at the two sides (fig. 203 B). In succeeding stages the contrast between the median and the lateral parts becomes still more marked, so that the epiblast becomes finally constituted of two lateral thickened bands, which meet in front in the procephalic lobes, and behind in the caudal lobe, and are elsewhere connected by a very thin layer (fig. 203 C). Shortly after the appendages begin to be formed, the first rudiments of the ventral nerve cord become established as epiblastic thickenings on the inner side of each of the lateral bands. The thickenings of the epiblast of the two sides are quite independent, as may be seen in fig. 203 C, _vn_, taken from a stage somewhat subsequent to their first appearance. They are developed from before backwards, but either from the first, or in any case very soon afterwards, cease to form uniform thickenings, but constitute a linear series of swellings--the future ganglia--connected by very short less prominent thickenings of the epiblast (fig. 200 C). The rudiments of the ventral nerve cord are for a long time continuous with the epiblast, but shortly after the establishment of the dorsal surface of the embryo they become separated from the epiblast and constitute two independent cords, the histological structure of which is the same as in other Tracheata (fig. 206, _vn_).

[FIG. 203. TRANSVERSE SECTIONS THROUGH THE VENTRAL PLATE OF AGELENA
LABYRINTHICA AT THREE STAGES.

A. Stage when about three segments are formed. The mesoblastic plate
is not divided into two bands.
B. Stage when six segments are present (fig. 200 B). The mesoblast
is now divided into two bands.
C. Stage represented in fig. 200 D. The ventral cords have begun to
be formed on thickenings of the epiblast, and the limbs are
established.
_ep._ epiblast; _me._ mesoblast; _me.s._ mesoblastic somite; _vn._
ventral nerve cord; _yk._ yolk.]

The ventral cords are at first composed of as many ganglia as there are segments. The foremost pair, belonging to the segment of the cheliceræ, lie immediately behind the stomodæum, and are as independent of each other as the remaining ganglia. Anteriorly they border on the supra-oesophageal ganglia. When the yolk-sack is formed in connection with the ventral flexure of the embryo, the two nerve cords become very widely separated (fig. 206, _vn_) in their middle region. At a later period, at the stage represented in fig. 201 B, they again become approximated in the ventral line, and delicate commissures are formed uniting the ganglia of the two sides, but there is no trace at this or any other period of a median invagination of epiblast between the two cords, such as Hatschek and other observers have attempted to establish for various Arthropoda and Chætopoda. At the stage represented in fig. 201 A the nerve ganglia are still present in the abdomen, though only about four ganglia can be distinguished. At a later stage these ganglia fuse into two continuous cords, united however by commissures corresponding with the previous ganglia.

The ganglia of the cheliceræ have, by the stage represented in fig. 201 B, completely fused with the supra-oesophageal ganglia and form part of the oesophageal commissure. The oesophageal commissure is however completed ventrally by the ganglia of the pedipalpi.

The supra-oesophageal ganglia are formed independently of the ventral cords as two thickenings of the procephalic lobes (fig. 205). The thickenings of the two lobes are independent, and each of them becomes early marked out by a semicircular groove (fig. 200 D) running outwards from the upper lip. Each thickening eventually becomes detached from the superficial epiblast, but before this takes place the two grooves become deeper, and on the separation of the ganglia from the epiblast, the cells lining the grooves become involuted and detached from the skin, and form an integral part of the supra-oesophageal ganglia.

At the stage represented in fig. 201 B the supra-oesophageal ganglia are completely detached from the epiblast, and are constituted of the following parts: (1) A dorsal section formed of two hemispherical lobes, mainly formed of the invaginated lining of the semicircular grooves. The original lumen of the groove is still present on the outer side of these lobes. (2) Two central masses, one for each ganglion, formed of punctiform tissue, and connected by a transverse commissure. (3) A ventral anterior lobe. (4) The original ganglia of the cheliceræ, which form the ventral parts of the ganglia[183].

[183] For further details _vide_ self, No. 440.

The later stages in the development of the nervous system have not been worked out.

The development of the nervous system in the Scorpion is almost identical with that in Spiders, but Metschnikoff believes, though without adducing satisfactory evidence, that the median integument between the two nerve cords assists in forming the ventral nerve cord. Grooves are present in the supra-oesophageal ganglia similar to those in Spiders.

The mesoblast. The history of the mesoblast, up to the formation of a ventral plate subjacent to the thickened plate of epiblast, has been already given. The ventral plate is shewn in fig. 203 A. It is seen to be formed mainly of small cells, but some large cells are shewn in the act of passing into it from the yolk. During a considerable section of the subsequent development the mesoblast is confined to the ventral plate.

[FIG. 204. LONGITUDINAL SECTION THROUGH AN EMBRYO OF AGELENA
LABYRINTHICA.

The section is through an embryo of the same age as that represented
in fig. 200 C, and is taken slightly to one side of the middle
line so as to shew the relation of the mesoblastic somites to the
limbs. In the interior are seen the yolk segments and their
nuclei.
1-16. the segments; _pr.l._ procephalic lobe; _do._ dorsal
integument.]

[FIG. 205. SECTION THROUGH THE PROCEPHALIC LOBES OF AN EMBRYO OF
AGELENA LABYRINTHICA.

The section is taken from an embryo of the same age as fig. 200 D.
_st._ stomodæum; _gr._ section through semicircular groove
in procephalic lobe; _ce.s._ cephalic section of body cavity.]

The first important change takes place when about six somites are established; the mesoblast then becomes divided into two lateral bands, shewn in section in fig. 203 B, which meet however in front in the procephalic lobes, and behind in the caudal lobes. Very shortly afterwards these bands become broken up into a number of parts corresponding to the segments, each of which soon becomes divided into two layers, which enclose a cavity between them (_vide_ fig. 204 and fig. 207). The outer layer (somatic) is thicker and attached to the epiblast, and the inner layer (splanchnic) is thinner and mainly, if not entirely, derived (in Agelena) from cells which originate in the yolk. These structures constitute the mesoblastic somites. In the appendage-bearing segments the somatic layer of each of them, together with a prolongation of the cavity, is continued into the appendage (fig. 203 C). Since the cavity of the mesoblastic somites is part of the body cavity, all the appendages contain prolongations of the body cavity. Not only is a pair of mesoblastic somites formed for each segment of the body, but also for the procephalic lobes (fig. 205). The mesoblastic somites for these lobes are established somewhat later than for the true segments, but only differ from them in the fact that the somites of the two sides are united by a median bridge of undivided mesoblast. The development of a somite for the procephalic lobes is similar to what has been described by Kleinenberg for Lumbricus (p. 339), but must not be necessarily supposed to indicate that the procephalic lobes form a segment equivalent to the segments of the trunk. They are rather equivalent to the præ-oral lobe of Chætopod larvæ. When the dorsal surface of the embryo is established a thick layer of mesoblast becomes formed below the epiblast. This layer is not however derived from an upgrowth of the mesoblast of the somites, but from cells which originate in the yolk. The first traces of the layer are seen in fig. 204, _do_, and it is fully established as a layer of large round cells in the stage shewn in fig. 206. This layer of cells is seen to be quite independent of the mesoblastic somites (_me.s_). The mesoblast of the dorsal surface becomes at the stage represented in fig. 201 B divided into splanchnic and somatic layers, and in the abdomen at any rate into somites continuous with those of the ventral part of the mesoblast. At the lines of junction of successive somites the splanchnic layer of mesoblast dips into the yolk, and forms a number of transverse septa, which do not reach the middle of the yolk, but leave a central part free, in which the mesenteron is subsequently formed. At the insertion of these septa there are developed widish spaces between the layers of somatic and splanchnic mesoblast, which form transversely directed channels passing from the heart outwards. They are probably venous. At a later stage the septa send out lateral offshoots, and divide the peripheral part of the abdominal cavity into a number of compartments filled with yolk. It is probable that the hepatic diverticula are eventually formed in these compartments.

[FIG. 206. TRANSVERSE SECTION THROUGH THE THORACIC REGION OF AN
EMBRYO OF AGELENA LABYRINTHICA.

The section is taken from an embryo of the same age as fig. 201 A,
and passes through the maximum protuberance of the ventral
yolk-sack.
_vn._ ventral nerve cord; _yk._ yolk; _me.s._ mesoblastic somite;
_ao._ aorta.]

The somatic layer of mesoblast is converted into the muscles, both of the limbs and trunk, the superficial connective tissue, nervous sheath, etc. It probably also gives rise to the three muscles attached to the suctorial apparatus of the oesophagus.

The heart and aorta are formed as a solid rod of cells of the dorsal mesoblast, before it is distinctly divided into splanchnic and somatic layers. Eventually the central cells of the heart become blood corpuscles, while its walls are constituted of an outer muscular and inner epithelioid layer. It becomes functional, and acquires its valves, arterial branches, etc., by the stage represented in fig. 201 B.

The history of the mesoblast, more especially of the mesoblastic somites, of the Scorpion is very similar to that in Spiders: their cavity is continued in the same way into the limbs. The general character of the somites in the tail is shewn in fig. 207. The caudal aorta is stated by Metschnikoff to be formed from part of the mesenteron, but this is too improbable to be accepted without further confirmation.

The hypoblast and alimentary tract. It has already been stated that the yolk is to be regarded as corresponding to the hypoblast of other types.

For a considerable period it is composed of the polygonal yolk cells already described and shewn in figs. 203, 204, and 205. The yolk cells divide and become somewhat smaller as development proceeds; but the main products of the division of the yolk nuclei and the protoplasm around them are undoubtedly cells which join the mesoblast (fig. 203 A). The permanent alimentary tract is formed of three sections, viz. stomodæum, proctodæum, and mesenteron. The stomodæum and proctodæum are both formed before the mesenteron. The stomodæum is formed as an epiblastic pit between the two procephalic lobes (figs. 200 and 205, _st_). It becomes deeper, and by the latest stage figured is a deep pit lined by a cuticle and ending blindly. To its hinder section, which forms the suctorial apparatus of the adult, three powerful muscles (a dorsal and two lateral) are attached.

[FIG. 207. TAIL OF AN ADVANCED EMBRYO OF THE SCORPION TO ILLUSTRATE
THE STRUCTURE OF THE MESOBLASTIC SOMITES. (After Metschnikoff.)

_al._ alimentary tract; _an.i._ anal invagination; _ep._ epiblast;
_me.s._ mesoblastic somite.]

The proctodæum is formed considerably later than the stomodæum. It is a comparatively shallow involution, which forms the rectum of the adult. It is dilated at its extremity, and two Malpighian vessels early grow out from it.

The mesenteron is formed _in the interior of the yolk_. Its walls are derived from the cellular elements of the yolk, and the first section to be formed is the hinder extremity, which appears as a short tube ending blindly behind in contact with the proctodæum, and open to the yolk in front. The later history of the mesenteron has not been followed, but it undoubtedly includes the whole of the abdominal section of the alimentary canal of the adult, except the rectum, and probably also the thoracic section. The later history of the yolk which encloses the mesenteron has not been satisfactorily studied, though it no doubt gives rise to the hepatic tubes, and probably also to the thoracic diverticula of the alimentary tract.

The general history of the alimentary tract in Scorpio is much the same as in Spiders. The hypoblast, the origin of which as mentioned above is somewhat uncertain, first appears on the ventral side and gradually spreads so as to envelop the yolk, and form the wall of the mesenteron, from which the liver is formed as a pair of lateral outgrowths. The proctodæum and stomodæum are both short, especially the former (_vide_ fig. 207).

_Summary and general conclusions._

The embryonic forms of Scorpio and Spiders are very similar, but in spite of the general similarity of Chelifer to Scorpio, the embryo of the former differs far more from that of Scorpio than the latter does from Spiders. This peculiarity is probably to be explained by the early period at which Chelifer is hatched; and though a more thorough investigation of this interesting form is much to be desired, it does not seem probable that its larva is a primitive type.

The larvæ of the Acarina with their peculiar ecdyses are to be regarded as much modified larval forms. It is not however easy to assign a meaning to the hexapodous stage through which they generally pass.

With reference to the segments and appendages, some interesting points are brought out by the embryological study of these forms.

The maximum number of segments is present in the Scorpion, in which nineteen segments (not including the præ-oral lobes, but including the telson) are developed. Of these the first twelve segments have traces of appendages, but the appendages of the six last of these (unless the pecten is an appendage) atrophy. In Spiders there are indications in the embryo of sixteen segments and in all the Arachnida, except the Acarina, at the least four segments bear appendages in the embryo which are without them in the adult. The morphological bearings of this fact are obvious.

It deserves to be noted that, in both Scorpio and the Spider, the cheliceræ are borne in the embryo by the first post-oral segment, and provided with a distinct ganglion, so that they cannot correspond (as they are usually supposed to do) with the antennæ of Insects, which are always developed on the præ-oral lobes, and never supplied by an independent ganglion.

The cheliceræ would seem probably to correspond with the mandibles of Insects, and the antennæ to be absent. In favour of this view is the fact that the embryonic ganglion of the mandibles of Insects is stated (cf. Lepidoptera, _Hatschek_, p. 340) to become, like the ganglion of the cheliceræ, converted into part of the oesophageal commissure.

If the above considerations are correct, the appendages of the Arachnida retain in many respects a very much more primitive condition than those of Insects. In the first place, both the cheliceræ and pedipalpi are much less differentiated than the mandibles and first pair of maxillæ with which they correspond. In the second place, the first pair of ambulatory limbs must be equivalent to the second pair of maxillæ of Insects, which, for reasons stated above, were probably originally ambulatory. It seems in fact a necessary deduction from the arguments stated that the ancestors of the present Insecta and Arachnida must have diverged from a common stem of the Tracheata at a time when the second pair of maxillæ were still ambulatory in function.

With reference to the order of the development of the appendages and segments, very considerable differences are noticeable in the different Arachnoid types. This fact alone appears to me to be sufficient to prove that the order of appearance of the appendages is often a matter of embryonic convenience, without any deep morphological significance. In Scorpio the segments develop successively, except perhaps the first post-oral, which is developed after some of the succeeded segments have been formed. In Spiders the segment of the cheliceræ, and probably also of the pedipalpi, appears later than the next three or four. In both these types the segments arise before the appendages, but the reverse appears to be the case in Chelifer. The permanent appendages, except the cheliceræ, appear simultaneously in Scorpions and Spiders. The second pair appears long before the others in Chelifer, then the third, next the first, and finally the three hindermost.

BIBLIOGRAPHY.

_Scorpionidæ._

(434) El. Metschnikoff. "Embryologie des Scorpions." _Zeit. f. wiss. Zool._ Bd. XXI. 1870.

(435) H. Rathke. _Reisebemerkungen aus Taurien_ (Scorpio). Leipzig, 1837.

_Pseudoscorpionidæ._

(436) El. Metschnikoff. "Entwicklungsgeschichte d. Chelifer." _Zeit. f. wiss. Zool._, Bd. XXI. 1870.

(437) A. Stecker. "Entwicklung der Chthonius-Eier im Mutterleibe und die Bildung des Blastoderms." _Sitzung. königl. böhmisch. Gesellschaft Wissensch._, 1876, 3. Heft, and _Annal. and Mag. Nat. History_, 1876, XVIII. 197.

_Phalangidæ._

(438) M. Balbiani. "Mémoire sur le développement des Phalangides." _Ann. Scien. Nat._ Series V. Vol. XVI. 1872.

_Araneina._

(439) M. Balbiani. "Mémoire sur le développement des Aranéides." _Ann. Scien. Nat_. Series V. Vol. XVII. 1873.

(440) F. M. Balfour. "Notes on the development of the Araneina." _Quart. Journ. of Micr. Science_, Vol. XX. 1880.

(441) J. Barrois. "Recherches s. l. développement des Araignées." _Journal de l'Anat. et de la Physiol._ 1878.

(442) E. Claparède. _Recherches s. l'évolution des Araignées._ Utrecht, 1862.

(443) Herold. _De generatione Araneorum in Ovo._ Marburg, 1824.

(444) H. Ludwig. "Ueber die Bildung des Blastoderms bei den Spinnen." _Zeit. f. wiss. Zool._, Vol. XXVI. 1876.

_Acarina._

(445) P. van Beneden. "Développement de l'Atax ypsilophora." _Acad. Bruxelles_, t. XXIV.

(446) Ed. Claparède. "Studien über Acarinen." _Zeit. f. wiss. Zool._, Bd. XVIII. 1868.

_Formation of the layers and the embryonic envelopes in the
Tracheata._

There is a striking constancy in the mode of formation of the layers throughout the group. In the first place the hypoblast is not formed by a process which can be reduced to invagination: in other words, there is no gastrula stage.

Efforts have been made to shew that the mesoblastic groove of Insects implies a modified gastrula, but since it is the essence of a gastrula that it should directly or indirectly give rise to the archenteron, the groove in question cannot fall under this category. Although the mesoblastic groove of Insects is not a gastrula, it is quite possible that it is the rudiment of a blastopore, the gastrula corresponding to which has now vanished from the development. It would thus be analogous to the primitive streak of Vertebrates[184].

[184] The primitive streak of Vertebrates, as will appear in the
sequel, has no connection with the medullary groove, and is the
rudiment of the blastopore.

The growth of the blastoderm over the yolk in Scorpions admits no doubt of being regarded as an epibolic gastrula. The blastopore would however be situated dorsally, a position which it does not occupy in any gastrula type so far dealt with. This fact, coupled with the consideration that the partial segmentation of Scorpio can be derived without difficulty from the ordinary Arachnidan type (_vide_ p. 120), seems to shew that there is no true epibolic invagination in the development of Scorpio.

On the formation of the blastoderm traces of two embryonic layers are established. The blastoderm itself is essentially the epiblast, while the central yolk is the hypoblast. The formation of the embryo commences in connection with a thickening of the blastoderm, known as the ventral plate. The mesoblast is formed as an unpaired plate split off from the epiblast of the ventral plate. This process takes place in at any rate two ways. In Insects a groove is formed, which becomes constricted off to form the mesoblastic plate: in Spiders there is a keel-like thickening of the blastoderm, which takes the place of the groove.

The unpaired mesoblastic plate becomes in all forms very soon divided into two _mesoblastic bands_.