The Apodidae; a morphological study

Front matter

f

-

A specimen of the Spitzbergen variety of Apus glacialis ( X 10), to show the general _ arrangement of the limbs and their gradual diminution from front to back. The white masses on each side near the brood pouches are partially extruded eggs.

Table Of Contents

Introduction xi

§ I. Object and Line of Argument i

§11. The Outer Body Form . . . • n

SECTION I Object and Line of Argument

The Apodidae have been known and studied for the last one hundred and fifty years. They have always attracted considerable attention, not only on account of their great size in comparison with other fresh-water Entomostraca, but also on account of their strange and sudden appearance in pools and ditches which owe their water entirely to the rainfall. This also is not all : their morphology has been a perpetual puzzle to zoologists, and they have been classed by some with archaic forms such as the Trilobites and Limulus, while by others they have been considered as highly specialised recent forms.

This essay claims by a new explanation of the morphology of the Crustacea, to set this latter point 3E B at rest, and to show that Apus must not only be ranked by the side of the Trilobites as one of the primitive Crustacean forms, but that it is itself a true link between the living Crustacea and the Annelida. By careful examination of the organisation of Apus, and a comparison of it with that of a carnivorous Annelid, it is possible to show, as will be done in the following pages, that Apus is perhaps the most perfect “ missing link ” which zoology so far possesses, perfect, not only because its morphology is easily deducible from that of a carnivorous Annelid, but also because the mechanical causes of the transformation are apparent. The Apodidae will in fact be found to afford us the first complete illustration of the rise of one large animal class out of another by the simple and natural adaptation on the part of one single species of the latter to a new manner of life. Close investigation shows the Apodidae to be both morphologically and biologically an almost ideal transition form.

More or less satisfactory transition forms between most of the great animal classes are now known, but none has till now been discovered between the Annelida and the Crustacea. The object of this book is to satisfy this want, not by the discovery of a new animal, but by a new explanation of one long known and often described.

The established transition forms between the other classes of the animal kingdom still leave much to be desired. Between the Protozoa and Metazoa the transition forms are either claimed, by botanists, or else, however probable, are somewhat hypothetical.

sect, i OBJECT AND LINE OF ARGUMENT

Between the Coelenterata and the Platodes we have rival links. When reading the arguments in favour of the claims of those specialised Ctenophora, the Cceloplana and Ctenoplana, we feel convinced ; but, on the other hand, when we study for ourselves under the microscope such a simple Rhabdocoele as Microstomum lineare, especially during its changes of shape when moving about under a cover glass, our former conviction fades away, and we see in it a specialised larval form of a Ccelenterate. Between the Platodes and the Annelids the gap seems small, but we cannot bridge it over until we decide whether the segmentation of the Annelids is a kind of axial strobilation, or the natural mechanical selection of internal symmetry. Between the Annelids and the Molluscs we have the claims of Solenogaster to attend to ; but this animal is unfortunately so rare, that it will be long before we can hope to have any very thorough knowledge of its morphology. ^Between the Annelids and the Tracheata we have Peripatus ; this highly interesting animal, has a special claim on our attention, as the Tracheata form with the Crustacea the great class known as the Arthropoda. We shall find that our explanation of the rise of the Crustacea supplies us also with a very probable clue as to the origin of the Tracheata. The Echinodermata and the Tunicata hover almost entirely in the air. And, lastly, we have the giant trunk of the Vertebrata, the roots of which are being eagerly sought in different directions. The claims of Amphioxus and of the Ascidian larva are confidently put forward by the majority of our leading zoologists, but there are difficulties not yet explained which make many restless, and lead them to search in other directions.

In this state of affairs it will be a clear gain and encouragement if we can connect the Annelida and the Crustacea in the way described in these pages, in which we show how a typical carnivorous Annelid (presumably a Nereid, though probably not so specialised as any modern member of that family) can, by a simple and natural adaptation to a new manner of life, be established as the ground type of Apus. We mean a great deal by this expression “ground type,” much more than any mere general resemblance of organisation ; we mean that every single organ of Apus, where it does not resemble that of its Annelid ancestor, is capable of being deduced from some organ in the latter, and, further, that the causes of the transformation are not far to seek. These are large claims ; the following pages will show whether they are justifiable.

Before entering into the morphological and anatomical details upon which our deduction of the Apodidae from a carnivorous Annelid is based, it will make the task of the reader lighter if we here set out the line of argument.

Many carnivorous Annelids have, as is well known, a protrusible pharynx, armed with teeth, which is shot out for the seizing of prey. We assume that the Annelid from which Apus is derived, adopted a habit of browsing, which rendered this protrusible

SECT. I OBJECT AND LINE OF ARGUMENT $ pharynx unnecessary, so that it degenerated. The Annelids afford us such a wonderful variety of forms adapted to almost every possible manner of life, that this assumption presents no difficulty. Cambrian and Silurian formations have revealed to the palaeontologist abundant evidence that early Chaetopods crawled about along the bottom of the seas of those times. That one of these should become specialised for feeding in the manner supposed, is not too much to ask.

The use of the pharynx just described is, as far as we can see, a clumsy method of obtaining food. The loss of it, and the adoption of a browsing method of feeding, might well be a gain. The further development of this habit would lead to a bending round of the head sufficient to enable the animal to use its anterior parapodia for pushing prey into its mouth. In time the bend of the head would become fixed, and the parapodia modified as jaws and maxillae. The parapodia of at least a certain number of anterior trunk segments would certainly also serve to rake food together into the middle line and forward it towards the mouth. From this very simple and natural modification of a Chaetopodan Annelid, we believe that all the Crustacea, living or extinct, can be deduced. To establish this, is the object of this little book, which we have called “ The Apodidae,” since it was during our study of these Phyllopods that we first caught sight of the Annelid, so effectually disguised under its Crustacean dress. Although this disguise is so complete as to have eluded all former research, yet when once under-

THE APODIDA:

stood, it is found to be very superficial. We shall be able to show, in the following pages, that the Apodidae agree in almost every detail of their organisation with such an Annelid, and that any disagreement is chiefly due to further specialisation in adaptation to the new manner of life described.

Commencing with the head, we shall show how the morphology of the typical Crustacean head is easily explained by the bending round of the five anterior segments of such an Annelid for the purpose of browsing.

The trunk of Apus will be shown to be a true link between the many-segmented Annelids, and the Crustacea with their small and almost constant number of segments. The rise of the shield will be briefly mentioned, a fuller account of it being reserved till we compare Apus with the Trilobites.

The gradual transformation of the Annelidan cuticle into the exoskeleton of the Crustacea, to which many of the changes in the inner organisation of the latter are to be referred, will be found well illustrated by the Apodidae.

The Annelidan parapodia (with their dorsal and ventral branches) will be shown to be capable of developing every form of Crustacean limb, the reasons for the suppression of one part and the development of another being generally fairly evident, Apus again supplying the clue.

Coming to the inner organisation, we shall take in turn, the musculature, the nervous system, the sensory organs, the alimentary canal, the circulatory system, sect, i OBJECT AND LINE OF ARGUMENT the excretory and other glands, and, lastly, the reproductive organs. We shall either point out the resemblances in each case between these orgaiTsTand those of our Annelid, or else show how they can be deduced from Annelidan organs. It will be found that while some of the modifications of Annelidan into Crustacean organs are easy to follow, the explanation of others has to be sought, and may thus appear to be, in some cases, far-fetched.

And here we must remind our readers that it is enough for our argument if we can show that such a deduction is possible. It is not essential to our theory that we should show exactly how the inner transformations actually took place. Our explanations may themselves be incorrect, but the validity of our argument can only be seriously weakened by showing that a set of organs in Apus could not possibly have been derived from any organs in the Annelida ; or that the improbability of such a transformation is so great that no experienced morphologist would accept it.

We shall conclude the first part of this essay by an appeal to the Nauplius, to see whether it bears out our theory that Apus is the original form of the majority of the modern Crustacea ; or, in other words, whether Apus can itself claim to be the proto-Nauplius of zoologists. We shall endeavour to describe the exact morphology of the Nauplius considered as the Apus larva or the Apus-stage in the development of the other Crustacea.

This will conclude Part I., which we hope will have shown that, so far as such claims can be based purely part r upon morphological, anatomical, and biological reasoning, the Apodidae deserve to take the place we assign them as an almost ideal transition form between the Annelida and the Crustacea. Here, as stated in the Preface, we thought to leave the matter as an interesting suggestion. Fortunately, however, we have the means of testing the accuracy of our conclusions..

Admitting, on the one hand, that the confirmatory evidence as to the truth of our theory given by the Nauplius need by no means be conclusive, we maintain, on the other hand, that the answer which we receive to our appeal to palaeontology and to such archaic living forms as Limulus must be decisive. Thus we enter upon the second part of our essay in order to obtain a final “ yea ” or “ nay ” as to whether our theory is, as a whole, but a morphological tour de force , or a fairly close guess at the truth.

We commence with Limulus, and show that if Apus is to be derived from an Annelid with the first five segments bent round ventrally, Limulus must have had a similar origin.

In the second section we venture into the dangerous realm of the Trilobites. The mystery which surrounds these primitive Crustacea is so great, that every , announcement of a new discovery bearing upon their morphology meets with more or less scepticism. Nevertheless, we believe that we can prove that our derivation of Apus from a bent Annelid reveals the Trilobites also in their true light, as so many attempts of browsing Crustacean-Annelids to adapt themselves to their surroundings — attempts which, in the long

SECT. I OBJECT AND LINE OF ARGUMENT run, proved unsuccessful, for reasons which we shall try to point out.

After briefly discussing the Eurypteridae, we shall give an outline sketch of a new classification of the Crustacea based upon our theory, showing that while only one group of modern Crustacea admits of derivation from the Trilobites, all the rest, except Limulus, can be deduced from the Apodidae. We shall see reasons for believing that it was the development of the shield, either as bivalve shell, or as a large fold of the tergum of the fifth segment, which led to success in the struggle for existence.

We should here say something as to the preservation of the Apodidae through so many geological ages. This is explained by the manner of life of the animals. They usually appear in ditches and pools dependent on the rainfall. In such waters they naturally come little into competition with other animals. The dry seasons are bridged over by the eggs being preserved in the mud. In this strange but perfectly natural way, Apus has, from the earliest times, been so completely isolated that its preservation presents no difficulty. Its presence in every part of the globe, with almost always the same manner of life, is a sign of its great antiquity.

The fact that no true fossil Apodidae are found, among the rich yield of Crustacean remains of the Silurian strata, admits of simple explanation. We say no true Apodidae, for we shall find that such forms as ITymenocaris and Ceratiocaris, though perhaps somewhat more specialised, were probably very closely

IO related to the Apodidae. Both these points will be discussed more in detail in Part II.

Finally, in a short concluding section we shall show that the method of differentiation which turned the Annelid into the Crustacean throws a flood of light on the origin of the Tracheata, and on some of the morphological differences which separate these two divisions of the Arthropoda.

Several new points in the anatomy of Apus will be described and illustrated. Where these do not bear directly upon the subject, they will be given in full in an appendix, so as not to interfere with the course of the argument.

Section II The Head

The comparative anatomy of the Crustacea has long ago established the fact that the Crustacean head must originally have been composed of five fused segments of an annulate body. Our derivation of Apus from a browsing Annelid explains the method of this fusion, that it did not take place along the longitudinal axis of the body, but by a doubling of this number of segments upon themselves. This term “ doubling ” or “ doublature ” has already been applied to the under sides of the forehead of such animals as Apus, Limulus, and the Trilobites, but apparently meaning nothing more than the doubling of the forehead, which has both a dorsal and a ventral surface. In reality, however, this “doubling” is the true description of the Crustacean head as shown in Figs, i and 2 ; these should be further compared with Fig. 46, p. 212, which represents a longitudinal section of a Trilobite, where the doubling is very clear.

Owing to the bending on itself of the cylindrical Annelidan body, the original head must have been anteriorly almost completely hemispherical (Figs. I and 4 6). This form of the original Crustacean-Annelid head

Fig. i. — Diagram showing the first six segments of a carnivorous Annelid ; the first five being bent round ; /, the prostomium with two pairs of eyes and a median cirrus, i, the ist segment carrying a pair of antennae, its under edge projecting backwards as the lower lip (/). 2, the end segment with a pair of antennal parapodia. 3, the 3rd segment with rudimentary dorsal parapodia, the ventral parapodia developing into mandibles. 4, the 4th segment with a pair of maxillae, the dorsal parapodium slightly less degenerated. 5, the 5th and last head segment, the dorsal parapodium with large aciculum and gland. 6, the 6th segment (ist free segment) with large dorsal parapodium carrying gill (f) and sensory cirrus ( c ).

was, however, very clearly modified. In most Trilobites traces of it are still visible in the glabella (Fig. 47, p. 213). In Apus, the disguise is very complete, the whole head being broad and flat. This form is due to a ridge running round the anterior surface as prolongation of the lateral edges of the shell fold. This ridge is of considerable interest, as it appears in almost every Trilobite. We shall later find reason to believe that it was a

Fig. 2. — Diagram of head and first trunk segment of Apus, for comparison with Fig. 1 — the lettering the same. In addition : s, shell gland (the acicular gland of Fig. 1 drawn into the shell fold). The distances between the limbs are much exaggerated in order to show their forms more clearly.

primary differentiation of the new “ head,” i.e. the new Crustacean head, composed of five Annelidan segments. Its origin and modifications will be discussed in another place. This form of head is no doubt useful for swimming and perhaps for burrowing in the mud.

Excepting in the number of limbs, all external traces of its having been composed of five segments are obliterated. Internally, however, there are abundant indications of its origin from the bending of the Annelidan segments.

The mouth lies ventrally and faces posteriorly, a fact which, taken in connection with the sharp bend in the oesophagus, to be described later, is enough of itself to suggest the original doubling of the segments. This ventral position of the mouth is an important characteristic of the whole class of the Crustacea, which has not received the attention it deserves.

Projecting ventrally and posteriorly over the mouth is a large upper lip, corresponding with the prostomium of the Annelid ancestor. This upper lip is thus a primitive feature among the Crustacea ; it occurs in a more or less pronounced form in most Nauplii, and persisted as a very highly developed organ in the Trilobites, but in the modern Crustacea it is generally more or less rudimentary

The under edge of the Annelidan mouth would also naturally project backwardsas a sharp fold (seeFig. I,/). Such a fold in Apus, however, would form an obstacle to the pushing of food forward towards and into the mouth by means of the maxillae and ventral parapodia of the anterior trunk segments ; hence we find it modified into two lateral projections, the middle part of the fold being merely indicated by a low ridge, which is not sufficient to form a barrier across the ventral surface. These two projections have been handed on to the higher Crustacea as the paragnatha.

In the larva of Euphausia they develop early as two limb-like projections- posterior to the mandibles, and strongly resemble, those of Apus.

In Apus, only the mandibles work between the labrum on the one hand and the under lip on the other. In Limulus, however, where the under lip also consists of two projections, the mouth is so stretched in the longitudinal direction that the masticatory ridges of five pairs of limbs work as jaws between them and the labrum. That these labial projections in Apus and Limulus are really homologous with one another, and with the under lip of our original Annelid, will be seen to follow as a necessary consequence of our explanation of their morphology.1

The shield is of great size, and stretches back (as a fold of the fifth segment) over the greater part of the body. Laterally it covers and protects the limbs. Posteriorly, it is armed with thorns, and has a keel along the dorsal middle line due no doubt to the central thorn which it supports. The carrying of these thorns, which are now so slightly developed, may have originally been the chief function of the shield in its early stages ; Fig. 48, page 215, in connection with which the origin of the shield will be discussed more in detail, illustrates what we imagine to have been the first step in the formation of the dorsal shield. This function has, however, long given place to that of protective covering of the whole dorsal surface and (laterally) of the gills. The lateral edges of the shield are prolonged into the ridge which, runi6 ning round the front of the head, makes the latter broad and flat, obliterates all external marks of segmentation, and effectually disguises its origin out of five Annelidan segments.

The coils of the shell glands form one of the most notable marks of the shield (see Frontispiece). Their origin, position, and structure will be discussed in the section on the excretory and other glands.

THE BODY PROPER.

On removing the shield we find a long vermiform annulate body, Fig. 3. In the anterior part of the trunk region the rings correspond in number with the limbs or parapodia ; as we approach the posterior region > however, the limbs are much more numerous than the rings. We find two, three, four, or as many as six rudimentary limbs on one ring. The last five rings have no limbs at all. This whole phenomenon, which has hitherto puzzled morphologists, may be explained as follows.

The great length of the original Annelid being of no use to the Crustacean-Annelid, the hinder part of the body remains in the latter at an undeveloped or larval stage. The rule in the development of Annelid larvae is that the successive segments form in front of the anal segment, and differentiate from before backward, those furthest from the anal segment being the most developed. In Apus, we find in front of the anal segment

1 7 five segments with no limbs developed, with no ventral ganglia, and with no organs except the most necessary, viz., the intestinal tube and the musculature. Then follows a row of rudimentary segments, each with a minute pair of limbs and a pair of ganglia, which increase in size and development from behind forward. The rudimentary segments which have become fixed in the adult Apus do bp

Fig. 3. — Lepidurus Spitzbergensis, from nature. The left half of the shield removed to show the vermiform body. The first 14 trunk segments carry a pair of limbs each, the following 10 “rings” carry between them ca. 28 rudimentary limbs, and therefore correspond to 28 segments. The last 5 trunk segments (excluding the anal segment) are limbless, bp, brood pouch formed by the nth pair of trunk limbs.

not correspond with the rings of the body ; only gradually as they recede from the limbless segments, and thus are more fully developed, do the segments correspond with the rings. The fusing of several rudimentary segments to form one body ring, i.e. a muscular segment, presents no difficulty. The Myriapoda afford us several examples of the fusing of true segments to form body rings.

The study of the development of Apus has shown

C

Part l that it grows directly out of the Nauplius by the gradual differentiation of new segments in front of the anal segment, with no metamorphosis worth mentioning. This fact has led to its being compared with an overgrown Nauplius. According to our view, indeed, the Nauplius is only the young Apus, or Apus-stage in other Crustaceans. This steady development of Apus from its larva, as an Annelid develops from the Trochophora, falls in with our explanation of its morphology. Apus, however, differs from its Annelid ancestors in that it reaches its adult shape, before its inherited number of segments are fully developed. This fixation of the hinder part of the body at a larval stage can be easily accounted for by the process of natural selection, compactness being a decided advantage to an active free-swimming animal.

The great number of segments, developed and rudimentary, in the Apodidae is a matter of considerable importance in estimating their true position. In all the other specialised Crustacean groups the number of segments is constant, i.e. constant for each group. In deriving such animals with a small but constant number of segments from Annelids with a large and varying number of segments, the ideal transition form would be an animal with a medium number of segments, which is not quite constant and is visibly diminishing. Both these points are specially clear in the Apodidse. We find that all descriptions of Apus cancriformis give a varying number of limbs, which can hardly be due merely to the difficulty of counting them, but more probably to the fact that the number of rudimentary limbs actually does vary. And even if it_ should be proved that the same species always possesses the same number of segments developed and rudimentary, the different species of Apus and Lepidurus are marked by decided differences in the number of segments. Whereas the more specialised Crustacea (the Malacostraca) have either the constant number of twenty or twenty-one segments, the number in the Apodidae varies between thirty-five and sixty-five. In the Entomostraca the number varies, but never reaches even the lowest number in the Apodidae. That the number in the Apodidae is visibly decreasing follows from our explanation of their morphology. The fact that the posterior segments remain fixed, in a larval and undeveloped condition, shows that they are gradually being dispensed with. On this ground alone, then, the Apodidae deserve to occupy the place, half way between the Annelids and the Crustacea, which we claim for them.

Many of the segments, as already seen, are so rudimentary as to be useless, i.e. as movable segments, so that three, four, or even six combine to form one body ring. In the Trilobites we shall find that the posterior rudimentary segments, which were for the same reason immovable upon one another, form ■ together, in many genera at least, a solid tail plate, the pygidium (cf. Fig. 49, p. 220).

It is a characteristic of the Crustacea that no limbs develop on the anal segment. In the Apodidae, this segment is already fully developed ; they are the

C 2

•segments anterior to it which are rudimentary, and which disappear in the development of the higher Crustacea.

The anal segment is provided with two long cercopoda or cirri, projecting posteriorly and slightly ventrally, and two rudiments, probably of similar appendages, on the posterior dorsal surface of the segment. These four together correspond with the four anal cirri found in some carnivorous Annelids {cf. pp. 85 and 274). The two cirri are stiffened for the greater part of their length by a thickened cuticle covered with setae, and showing slight rings of thinner skin. The tips of the cirri are quite thin-skinned, and seem to function as tactile papillae.

The posterior dorsal surface of the anal segment is sometimes prolonged into a variously shaped caudal plate or lamella,1 which we shall find to be the homologue of the caudal spine of the Xiphosuridae.

THE CUTICLE AND EXOSKELETON.

The generally thin and flexible Annelidan cuticle of the Apodidae shows local thickenings which may well be recognised as the commencement of the Crustacean exoskeleton. A closer study of these reveals to us the principles of the original formation

1 Apodidae having this characteristic have been classed by Leach as a separate genus, Lepidurus. But Dr. Alfred Walter, to whose memory this essay is dedicated, discovered a form in a desert well in Transcaspian Russia, Apus Hackelii (Walter), which makes it doubtful whether this division can be sustained. ( Bulletin de la Society ImpSriale des Naturalistes de Moscou, 1887.)

of such an exoskeleton. Perhaps the best way to make the subject clear is to discuss and illustrate these principles.

(1) First and chiefly, the cuticle is thickened for the protection of exposed parts. We find the cuticle of the dorsal surface of the head, which, on the bending round of the anterior segments, was left entirely unprotected.,- and of the upper surface of the shield, considerably thickened ; also that of the exposed segments, i.e. of those segments which are not covered by the dorsal shield. Underneath the shield, on the contrary, the skin of the body is very thin and flexible, though towards the posterior edge of the shield it begins to thicken. We find the same principle in Limulus and the Trilobites, where, under the protection of the thick shield, the cuticle of the ventral and lateral parts remained soft and flexible.

We also find certain parts of the body thickened for protection against other parts. Thus the outer edges of the under lips are thickened for protection against the working of the powerful jaws and of the first maxillae, between which two pairs of limbs they are placed.

(2) We find local thickenings to counteract the pulls of the muscles, and this in two ways, (a.) There are thickened areas such as the ventral and lateral parts of the mandibles, to resist the almost rectangular pulls of the mandibular muscles. In the higher Crustacea such thickenings of the cuticle go hand in hand with the concentration and physiological perfection of the muscle bands, which, instead of being attached to large irregular areas of a soft cuticle, as in the Annelids and in the trunk segments of Apus, are * attached to definite firm points. (A) There are thickened strips to resist the longitudinal pulls of muscles, as along the shafts of the limbs, where, but for such bands, the cuticle of the limb would be drawn into folds (Fig. 4).