The Eurypterida of New York/Volume 1/Taxonomic relations

by John Mason Clarke and Rudolf Ruedemann

VI
TAXONOMIC RELATIONS

Any inquiry into the zoological position of the eurypterids must be based on that of Limulus since there is little doubt in the minds of paleontologists of the close relationship between the eurypterids and the king crab. In fact, all recent investigations and discoveries of eurypterids have only served to bring out new homologies of structure between the two groups. Nieszkowski, Hall and Woodward among the earlier writers on eurypterids clearly recognized and distinctly pointed out the numerous homologies and united the eurypterids and xiphosurans. They were especially successful in correlating the segments and the appendages of the cephalothorax. Later Schmidt and Laurie established the fact that the Eurypterida had four pairs of platelike abdominal appendages that bear the branchiae on their posterior surfaces and constitute another important homology with Limulus. Finally Laurie and Holm indicated as a further common structure the existence of preoral appendages in the eurypterids. Holm especially demonstrated this close relationship with the limulids, by the description of a number of finer details of organization: the occurrence of epicoxae in the second to fourth endognathites, the presence of a circular hole spanned by a thin membrane in the coxa of the fourth endognathite. We believe we have added to this array of evidence some further important details in identity of structure of the compound eye of Pterygotus and that of Limulus, and of the general parallelism in the ontogeny of the eurypterids and Limulus. The former argument is considered of especial significance as the eye of Limulus has been shown by Watase and others to be of a type of structure entirely peculiar. In the chapter on ontogeny we have not expressly pointed out the similarities and differences in the individual development of the eurypterids and limulids, and may here briefly enumerate them. The nepionic stages of both have in common:

1. The large size of the carapace
2. Its broad border
3. Its distinct median, glabellalike ridge (best seen in Stylonurus myops)
4. The smaller number of body segments (nine or less, seen also in S. myops)
5. The lack of differentiation of the segments
6. The undeveloped telson

As differences appear:

1. The terete or conical abdomen in the young eurypterids in contrast with the broadened abdomen of Limulus, and
2. Large larval eyes in the eurypterids

We consider both these differential characters as due to purely adaptive changes. The broader abdomen of the larvae of Limulus results from the earlier appearance of the broad abdomen of the mature type through acceleration, and the adaptive nature of the large larval eyes has been fully discussed on page 119. For the reasons here given we find ourselves in agreement with those authors who have united the eurypterids and limulids under Dana's subclass Merostomata.

If the relationship of the eurypterids with the king crabs is so close that it places them in one subclass, the eurypterids will have to follow the Limulidae in their wanderings in the zoological system.

The limulids were, as is well known, classed with the crustaceans, chiefly on account of their aquatic habit and branchial respiration, since the Arthropod phylum was by common consent divided into two subphyla, the Branchiata and Tracheata. Opposition arose to this classification through the recognition of the affinities of Limulus with the arachnids, first suggested by Strauss-Durkheim, and especially elaborated by Van Beneden, Lankester, Kingsley and Laurie, though not without opposition from such authorities as Packard, Woodward, Thorell and Lindström, Bernard and others. Some, as Kingsley, have placed the Merostomata next to the Arachnida and united the two in one class. The scheme proposed by Kingsley [1894, p. 122] is as follows:

Phylum, Arthropoda

Subphylum I, Branchiata

Class I, Crustacea

Subclass 1, Trilobitae

Subclass 2, Eucrustaceae

Class II, Acerata

Subclass 1, Gigantostraca (Merostomata)

Subclass 2, Arachnida

Subphylum II, Insecta

Lankester, in his paper "Limulus an Arachnid?" and his followers have gone farther and placed the Merostomata under the class Arachnida. They propose the following classification [see Shipley, 1909, p. 258]:

Class, Arachnida

Subclass 1, Delobranchiata (Merostomata)

Order (I), Xiphosura

Order (II), Eurypterida

Subclass 2, Embolobranchiata

Order (I), Scorpionidae

Order (II), Pedipalpi

etc.

The Delobranchiata, which term is equivalent to Dana's Merostomata, are characterized by their gills which are patent and exposed; the Embolobranchiata have lungbooks or tracheae. We prefer to retain the term Merostomata, there being no apparent need for a new word.

As there is a burning interest attaching to the question whether we should regard the merostomes of the Siluric as giant marine arachnids or archaic crustaceans, we here briefly review the arguments for the relationship of the merostomes with the scorpions; and since, the larval stages of the eurypterids are described in detail now for the first time, it appears very appropriate to test the hypothesis of this arachnidan relationship by the ontogenetic evidence now available. Kingsley [1893, p. 228] cites, in his full discussion of the agreements and differences between Limulus and the crustaceans on one hand and the arachnids on the other, these six points of agreement:

1. A branchial respiration
2. The possession of biramous appendages
3. The absence of malpighian tubes
4. The absence of salivary glands
5. The absence of embryonic envelops
6. The presence of compound eyes

and 28 points in which Limulus and the arachnids agree, and in which both differ from the other "Tracheates" (Hexapoda and Myriapoda). The following of these points are considered as of special importance for the association of Merostomata and Arachnida:

(1) The numerical homologies of segments and appendages; (2) the exact homologies existing in the respiratory organs; (3) the fact that the cephalothoracic appendages are pediform, the basal joints serving as jaws; (4) the presence of true nephridia opening in the base of the third or fifth pair of appendages or in both; (5) genital openings in the seventh (or more probably eighth) segment of the body; (6) extreme length of the midgut; (7) presence of an internal structure, the entosternite; (8) inclusion of the ventral nerve cord and its nerves in the external artery and its branches; (9) the close similarities in the central nervous system.

The last chapter of Thorell's paper on A Silurian Scorpion from Gotland is entitled: Are the Merostomes Arachnids? Here the author gives a critical discussion of these points of resemblance and concludes that the characters on the basis of which the merostomes have been considered as nearly related to the scorpions are "fully counterbalanced by the differences exhibited in the structure of the organs of respiration and by the fact that the scorpions and all the higher Arachnida are provided with malpighian vessels, whereas the merostomes (Limulus) and all unquestionable Crustacea are devoid of these vessels." He suggests that the arachnids also form an ancient type, and that the common origin of the merostomes and the arachnids must be sought for far down in the pedigree of the arthropods. "The agreement between scorpions and eurypterids would then be derived from causes quite independent of close relationship, and perhaps from a convergence in some branches of the two stems which are formed by the merostomes and the arachnids."

Opposed to this, Lankester and his followers hold that the presence of gills and the absence of malpighian tubules are features associated with aquatic life, and hence of no critical classificatory value, and Laurie and Claus assert that the morphological value of the gills has been greatly overestimated. Laurie notes that the branchiae of eurypterids, like those of Limulus, are constructed on a type unknown amongst the Crustacea, and further, that structures such as these, which are the product of a physiological necessity, are not of much value in indicating close relationship. He cites as arguments against the crustacean relationship "the segmentation of the body and position of the genital aperture—which does not agree with that of any known crustacean, the absence of anything representing the first antennae, the chelate structure of the one pair of preoral appendages and the fact that there is no trace of the typical crustacean biramous structure of the appendages." Admitting the relationship of the eurypterids to the arachnids, Laurie thinks that "the eurypterids must be considered as intermediate between Crustacea and Arachnida, in the sense that they are among the most primitive arachnids, and therefore nearer the junction point of the two terms." Agreeing with the other authors as to the close relationship of the eurypterids with Limulus, Laurie sees a point of great morphological importance "in the fact that Limulus has a pair of platelike appendages on the second abdominal segment," and takes this to indicate that "Limulus branched off from the eurypterid stem before the genital operculum was so highly specialized as it is in the eurypterids, and, consequently, before the appendage of the second abdominal segment had become reduced." From this fact, and the absence in Limulus of anything comparable to the central lobe of the genital operculum, the higher state of development of the metastoma in eurypterids, and the higher specialization of the last pair of legs, it is inferred, that "Limulus represents a more primitive type in almost every respect except the fusion of the abdominal segments, and is to be related to the eurypterids
Figure 27 Protolimulus eriensis Williams. Chemung group, Erie county, Pennsylvania. (From Hall and Clarke) not by direct descent, but through a comparatively unspecialized ancestor."

Our evidence is fully accordant with Laurie's view of the early separation of the eurypterids and Limulus. Of especial importance in this regard is the occurrence of a typical eurypterid in the Cambric era (Strabops). The ventral side of this form is still unknown and in view of its important bearing on the phylogenetic relations of the eurypterids with the limulids and scorpions, the elucidation of the character of its genital operculum is especially desirable. At any rate, none of the eurypterids now known from the Cambric to the Permic even suggest any fusion of the abdominal segments such as distinguishes Limulus, while in the Devonic Protolumulus eriensis Williams, the abdominal segments are apparently already fused and the characteristic Limulus structure developed.

With the conclusion that the limulids and eurypterids were probably separate in Precambric time, the question arises as to the date of separation of the Scorpionidae from the eurypterids. The relationship of Scorpio to Limulus has been fully discussed, especially by Lankester, and is no longer doubted. That of Scorpio to the eurypterids has been more fully dealt with by Laurie. Lankester had already indicated the close homologies existing in the number of body segments of the cephalothoracic appendages [see especially op. cit. tabular statements p. 536].^[1] From these it would follow that Limulus on the one hand, Scorpio and the eurypterids on the other, separated before the consolidation of the body segments observable in Limulus; but Laurie states that the second abdominal segment in Scorpio is well developed and shows no sign of ever having been suppressed by the genital operculum as in the eurypterids. From this argument only it might seem that the scorpions came off the eurypterid stem before the great development of the genital operculum. Laurie considers this development of the genital operculum at the expense of the second free segment as a point of considerable morphological importance and has therefore [op. cit. 526, see also Recent additions, etc. p. 127] considered it probable that the Pedipalpi (Thelyphonus) are more nearly allied to the eurypterids than are the scorpions, for in the former a similar suppression of the second ventral segment has taken place in favor of the genital plate. The four scorpions known from the Siluric all exhibit characteristic features of their order. Recently a fossil Thelyphonus from the Bohemian Carbonic, Prothelyphonus bohemicus (Kusta), has been fully made known by Fritsch [1904]. Unfortunately the specimen shows the ventral aspect only, more or less involved with the dorsal side. But if we read the figures 5, 6 and 7 of his plate 6 aright, the ventral segments in Carbonic time still possessed a like development with the dorsal ones and the agreement of the Pedipalpi with the eurypterids
Figure 28 B, fourth leg of a recent scorpion (Buthus australis); C, third leg of siluric scorpion (Palaeophonus nuncius). (From Pocock) in the greater development of the genital plate would have manifested itself only after the disappearance of the eurypterids and so lack phylogenetic significance.

Some very significant points regarding the relationship of the eurypterids and scorpions have been brought out by Pocock's investigations [1901] of the Scottish Siluric scorpion, Palaeophonus hunteri. We briefly note the more important of these. Pocock points out that the walking legs of Palaeophonus differ from those of all other scorpions, living or fossil, in their primitive character [text fig. 28]. They consist of the primitive number of segments (seven), show the simplicity of segmentation by the subequality of the individual segments and possess a sharply pointed, practically clawless terminal segment, strikingly resembling those of some eurypterids (Pterygotus) and differing greatly from those of the recent scorpions. Further, in distinction from the later scorpions, the basal or coxal segments of all the appendages were in contact or capable of meeting in the middle line, although the coxae of the fourth were small and functionless [text fig. 85]. In this feature the archaic Palaeophonus presents a condition intermediate between that of the typical scorpions and of Limulus or Eurypterus. Likewise, it is shown that the comb is structurally intermediate between a typical scorpion comb and the outer branch of one of the metasomatic appendages of Limulus. Finally Pocock doubts that Palaeophonus possessed spiracles or was an air breather but suggests that it was still aquatic and may have possessed branchial lamellae attached to sternites.^[2]

We have above pointed out the features which the larvae of Limulus and the eurypterids have in common. The supposed close relationship of the eurypterids to Scorpio, makes it desirable to compare the larval stages of the two.

The embryology of the scorpions has been investigated by Metschnikoff and more recently by Laurie [1890] and Brauer [1895]. We copy here for comparison one of Metschnikoff's [from Balfour's Treatise on Comparative Embryology] and one of Brauer's figures. Study of these shows that the carapaces of the larval eurypterids have a closer agreement with
Figure 29 Early stage in development of scorpion. Ventral view of embryo. af, eye fold; sa, lateral eyes; obl, upper lip; Km, comb (pecten); abp, abdominal legs (gills). (From Brauer) Limulus than with the scorpions. In the scorpion embryo and larvae the carapace is relatively of larger size than in the eurypterids, being nearly half the length of the embryo, but it is long and slender, instead of short and broad or semicircular as in the merostomes; the procephalic region early, exhibits a deep frontal emargination and, according to Metschnikoff, a distinct bilobation, features that are found in neither Limulus nor the eurypterids, but which are characters of the adult scorpion. Likewise the development of the cephalothoracic appendages, while homologous in number and original position as to the stomodaeum (embryonic mouth) to that of the merostomes, differs from that of the eurypterids and Limulus in the early development of the pedipalps of the second segment, corresponding to their prominence in the adult scorpion. They are even recognizable earlier than the chelicerae of the first segment, while the following four pairs of walking legs are of uniform size and character.

The homology of the appendages of the cephalothorax repeats itself in the abdomen; here again the number is absolutely in agreement with that of the merostomes, but there appears to exist from the beginning a striking difference in the strong differentiation of the preabdomen and the postabdomen early in the embryonic life of the scorpion, while in the eurypterids that difference between the two parts of the abdomen is still obscure in the larva, as it is in Limulus. In the scorpion the preabdomen is broad and bulky while the postabdomen or tail is abruptly set off, very narrow and flexed upon the ventral surface [see text fig. 30].

Figure 30 Later embryonic stage in development of scorpion. Ventral view. ch, cherlicerae; p¹-p⁴, walking legs; pe, pecten; st, stigmata; ab, postabdomen. (From Balfour) The abdominal appendages of the embryo of the scorpion [see text fig. 29] atrophy except for the first pair which form the "combs." In the places of the others the lungbooks appear, resulting from paired invaginations, the walls of which subsequently become plicated. In the merostomes, on the other hand, the preabdominal appendages remain throughout life and bear the branchial lamellae. It is the current opinion of zoologists that the lungbooks of the scorpion are derived from such branchiate abdominal appendages as those of Limulus, the conversion of one set of organs into the other being supposed to have been effected by the formation, behind each pair of abdominal appendages, of an invagination which, deepening, has carried in with it the branchial lamellae. Brauer [1895, p. 373] has recognized the origin of the lungbooks from gills.

In summing up the comparison between the embryonic and larval stages of the scorpion and the larvae of the eurypterids we may say that, (1) the general homologies of the two are very apparent in the composition of the carapace and abdomen of an equal number of segments, but that, (2) while in the scorpion the segmentation is completed long before the hatching, in the eurypterids the larvae in the nepionic stage still lack the full complement of segments, recalling the trilobites in this feature and clearly representing a more primitive condition; (3) there are a number of distinct differences in the larvae of the eurypterids and of the scorpions, some of which lie in the form of the carapace, the early embryonic differentiation of the pedipalps from the other appendages, the strong prelarval differentiation of preabdomen and postabdomen and the disappearance of the abdominal appendages in the scorpions and their persistence in the eurypterids.

It is yet to be determined whether these differences are of phylogenetic significance or only due to the pushing back, by acceleration, of later adaptations of the scorpions into the embryonic and larval stages. The latter is undoubtedly the case in the embryonic development of the pedipalps and of the narrow, distinctly defined, scorpionic tail. The early appearance and later atrophy of the abdominal appendages is, however, clearly a feature that points to a common ancestor for the scorpion and the eurypterids having such appendages, and we believe that the cephalothorax in the embryo of the scorpion, retains ancestral features from the facts that its length corresponds to about six abdominal segments and it equals the latter in width [text fig. 29]; that, however, the strong development of the procephalic region is tachygenetic.

A comparison of the larvae of all three, the eurypterids, Limulus and the scorpion, shows both the latter to have lost the primitive form of the abdomen by acceleration, that of Limulus being much broadened, that of the scorpion abruptly contracted to the tail or postabdomen while the eurypterids have best preserved the original gradual and uniform contraction. The carapaces of the eurypterids and the scorpion have most nearly retained the original proportions and form of the common ancestor. Of the cephalothoracic appendages the chelicerae are alike in all three groups and obviously ancestral in their form; the remaining legs have taken quite different courses of adaptation, the scorpions having developed the powerful chelate pedipalps, the eurypterids the swimming legs, while those of Limulus have remained relatively undifferentiated, and show no tachygenetic features in the embryos except the chelae. The embryo of the scorpion shows simple walking legs, like those of the eurypterids, and lacks the two movable claws. This simple form of the walking leg is also exhibited in a very remarkable manner by the Siluric Palaeophonus. It is quite safe to infer that this is the form of the cephalothoracic appendages inherited from the common ancestor.

A corollary of these inferences is that neither Limulus nor the scorpions is derivable from the eurypterids, but that all three, while related, have early separated; and that the eurypterids are still nearest in their general aspect to this common ancestor. The early authors in pointing to the "larval aspect" of the eurypterids, showed therefore, a very clear insight into the phylogenetic status of this subclass. The appearance of the eurypterids in the Cambric with the essential characters of the group is in accordance with this larval aspect while the early separation of the scorpions from the stock is evinced by the occurrence of typical scorpions in the Siluric, and by the fact that in the Carbonic they show a greater diversity of form than they do today. On the other hand the similarity of the Palaeophonus nuncius to recent forms is conclusive evidence that the scorpions have been very "persistent types" and have carried their typical characters well back of the Siluric. There is no reason to doubt that, as there are eurypterids in the Cambric, the scorpions also reach back to that era and the diversion from the common ancestor must have already been inaugurated in early Cambric time.

As to what this common ancestor was we have no clue. The trilobites were commonly adduced as competent to furnish it, they, the Xiphosura and eurypterids, having been united as "Poecilopoda," until the phyllopodiform structure of the trilobite limbs was demonstrated by Beecher and they were recognized as true, primitive Crustacea. While the trilobites are separated by a series of features that effectively characterize them as primitive Crustacea (as the protonauplius, the hypostoma, the slender jointed antennules, the biramous character of all other limbs, the compound eyes on free cheek pieces, etc.) and that disprove any assumption of their ancestral relations to the Merostomata, phylogenists still assert that the resemblances between the crustaceans and the Acerata (Merostomata and Arachnida) are much closer than those between either and any of the other groups of arthropods and they would derive the Arachnida through the Merostomata from primitive crustaceans. Laurie [Recent Additions, etc. p. 117] however, has pointed out that the relationship of the eurypterids to the Crustacea is not so evident, as they show no special points of affinity with any one group. "The absence of that special modification of three pairs of appendages to serve as mouth organs, which is characteristic of all Crustacea except the Ostracoda, indicates that their point of union must have been very low down the crustacean stem, and the very definite number of segments and arrangements of appendages in the Eurypteridae indicates on the other hand that they are removed a considerable distance from any such primitive type." The fixation of the number of segments in the Cambric eurypterid Strabops is significant as indicating that at this early date they were already far removed from the common ancestor with an indefinite number of segments.

As the most primitive and earliest crustaceans, the trilobites, are clearly not ancestrally or otherwise closely related to the eurypterids and the latter even in the Cambric are far removed from any possible synthetic ancestors, it is a fair question whether it is not proper to look for more primitive arthropods than the crustaceans as ancestors of the eurypterids. We have in mind now the investigations of Bernard [1896] who disputes the relationship of the merostomes to the crustaceans on one hand and to the arachnids on the other, and states that "As arthropods, no relation whatever exists between them; as segmented animals, however, they are both derivatives from the chaetopod annelids, but along different and opposite lines of specialization." Bernard derives the Crustacea from a bent carnivorous annelid, a view which Beecher regards as partly verified by his discoveries concerning the ventral anatomy of the trilobites, and it is therefore worthy of consideration in this place. If we consider the absence of anything in the ontogeny of the eurypterids that would suggest a crustacean nauplius stage, the admitted absence of all crustacean features in the adult forms, and the equal absence of all crustacean features in the ontogenies of Limulus and the scorpion, it becomes manifest that the inference of the derivation of these classes from arthropods more primitive than the Crustaceans, seems well founded.

Through the suggestive investigations of Professor Patten, published in his paper On the Origin of Vertebrates from Arachnids and the ingenious speculations of Dr Gaskell as to the origin of the vertebrates from some Limuluslike ancestor, the eurypterids have gained increased interest and some notoriety as possible ancestors of the earliest vertebrates.

---

1. A brief summary of the resemblance of the eurypterids with the scorpion has been lately given by Woods [1909, p. 283]. It reads as follows:
   The eurypterids present a striking resemblance to scorpions. In both groups the segments in the three regions of the body are the same in number, and the appendages of the prosoma also agree in number and position. The preoral appendages are chelate in both, but the second pair of appendages are chelate in the scorpions only. In eurypterids the coxae of the five pairs of legs are toothed and meet in the middle line, but in the scorpions the coxae of the last two pairs do not meet; this difference, however, appears to be bridged over in the earliest known scorpion—Palaeophonus, from the Silurian rocks. The eurypterids are distinguished from the scorpions by the much greater development of the last pair of legs. The large metastoma of the former is homologous with the sternum of the scorpion. The genital operculum is much smaller in scorpions than in eurypterids, and in this respect the latter agree with Thelyphonus (one of the Pedipalpi) more than with the scorpions. The pectines are absent in the eurypterids except in Glyptoscorpius. Instead of the lung books of the scorpions the eurypterids possess branchial lamellae on the platelike appendages; but this difference between the two groups appears to be bridged over by Palaeophonus, which was marine, and may have possessed branchial lamellae since stigmata seem to be absent.
2. His arguments are that the Scottish Palaeophonus hunteri does not show the stigmata, which Peach believed he saw, and that the single stigma seen by Thorell and Lindstrom in P. nuncius is a fortuitous crack. He therefore holds that Palaeophonus had no stigmata and spiracles and that on account of the excellent preservation of the Siluric scorpions in undoubted marine beds, they can not have been land animals, and that the strong sharply pointed legs were admirably fitted, like those of a crab, for maintaining a secure hold amongst the seaweed.
   In regard to the supposed absence of stigmata in P. hunteri, notwithstanding the fact that the ventral side is exposed, it may be mentioned that Fritsch [1904, p. 64] has pointed out that the relation of the chelicerae to the frontal margin of the carapace shows that the specimen (of P. hunteri) lies with the dorsal side up and with the ventral organs of the cephalothorax pressed through the mutilated carapace. In that case it can not be expected that the fine slitlike stigmata should be observable, and in all Siluric scorpions which happen to have only the dorsal sides exposed, the question of the presence or absence of stigmata is obviously still an open one. Nevertheless Pocock's view of the aquatic habit of Palaeophonus is of interest in connection with the New York Proscorpius osborni in view of the absence of all other remains of land animals or plants in the waterlime; and especially in view of Brauer's discovery [1895, p. 351] that the ontogeny of Scorpio shows that the lungbooks are derived from gills borne on mesosomatic appendages.