The Cockroach of the Past.

By S.H. SCUDDER, of the U.S. Geological Survey.

Like all useful scavengers, the Cockroach is looked upon nowadays as an unmitigated pest. It has, however, a certain right to our regard, for it comes of a venerable antiquity. Indeed, palÆontologically considered, no Insect is so interesting as the Cockroach. Of no other type of Insects can it be said that it occurs at every horizon where Insects have been found in any numbers; in no group whatever can the changes wrought by time be so carefully and completely studied as here; none other has furnished more important evidence concerning the phylogeny of Insects. Even the oldest known air-breathing animal has been claimed (though I think erroneously) as a Cockroach; yet, however that may be, it is certain that in the most ancient deposits which have yielded any abundance of Insect remains, the Coal Measures, they so far outnumber all other types of Insects, that this period, as far as its hexapodal fauna is concerned, may fairly be called the Age of Cockroaches. And though the subsequent periods show an ever-diminishing percentage of this family when compared with the total synchronous Insect fauna, yet the existing species are counted by hundreds, and the fecundity of some, attested by every housewife, may be looked upon as a sufficient explanation of the persistence of this antique type. The Cockroach is, therefore, a very aristocrat among Insects.

Our knowledge of its past is derived almost entirely from its wings; perhaps because these organs are the farthest removed from the nourishing fluids of the body, which on death become one of the agents, or at least the media, of putrefaction and consequent obliteration. At all events, whatever the cause, these chitinous membranes, with their network of supporting rods, and even not infrequently with the minutest reticulation of the membrane itself, are preserved with extraordinary fidelity, and in such abundance that, by comparison with similar parts in existing forms, we may reach some general conclusions concerning the life of the past of no little interest.

The first thing that would strike an observer, looking at the ancient Cockroaches, would be their general resemblance to the living. Excepting for their usually larger size,194 were we to have the oldest known Cockroaches in our kitchens to-day, the householder would take no special note of them—unless, indeed, the transparency of their wings (shortly to be mentioned) were to give them a somewhat peculiar aspect. There would be the same rounded pronotal shield, the same overlapping wings, coursed by branching veins, the same smooth curves and oval flattened form of the whole creature, and doubtless also the same scurrying movements. Indeed, some accurate observers—so, I suppose, we must call them—have failed to take note of some important and very general distinctions between the living and the dead. Thus Gerstaecker, in a work begun twenty years ago, and not yet finished, said, near its beginning,195 “Not a single species of Insect has yet been found in the Carboniferous rocks which does not fall, on closer examination (mit voller Evidenz), not only in an existing order, but even almost completely in the same family as some living form, and only presents striking distinctions when compared with the species themselves.” He further specifies the Cockroaches described from the Coal Measures, by Germar and Goldenberg, as agreeing in every distinguishing family characteristic with those of the present day.

In one sense, indeed, this is true. We separate the living Cockroaches from other kinds of Orthoptera as a “family” group, and “Cockroaches” have existed since the Coal Measures at least; yet the structure of every one of the older types is really so peculiar that none of them can be brought within the limits of the family as it now exists. We recognise ours, indeed, as the direct descendants of the ancient forms, but so changed in structure as to form a distinct group. A parallel case is found in the Walking-sticks, and is even more obvious. The recent researches of M. Charles Brongniart have brought to view a whole series of forms in Carboniferous times, which are manifestly the progenitors of living Walking-sticks, with their remarkably long and slender stick-like body, attenuated legs, and peculiar appendages at the tip of the abdomen. Existing forms are either wingless or else have opaque elytron-like front wings, and very ample, gauzy, fan-like hind wings; while the Carboniferous species are furnished with four membranous wings, almost precisely alike, and so utterly different from those of existing types that, before the discovery of the bodies, these wings were universally classed as the wings of Neuropterous Insects (sensu Linneano). Thus Gerstaecker, in the very place already quoted, says of these same wings, known under the generic name Dictyoneura, that they show at least a very close relationship to the EphemeridÆ of to-day.

One principal difference here alluded to—the exact resemblance, except in minor details, of the front and hind wings, and, as consequent therewith, equal diaphaneity in both—is found indeed in all palÆozoic insects, with exceedingly few exceptions;196 it is one of their most characteristic and pervading peculiarities. It marks one phase of the movement in all life from homogeneity to heterogeneity—from the uniform to the diverse. In the Cockroaches of to-day a few are found in which the tegmina are nearly as diaphanous as the hind wings; but in the great mass the texture of the tegmina, as in Orthoptera generally (excepting most Gryllides), is decidedly coriaceous; and in some, e.g., Phoraspis, the veins are nearly obliterated in the thickness and opacity of the membrane, so as to resemble many Coleopterous elytra.

Three principal differences have been noticed between the ancient and modern forms of Cockroaches. Doubtless others could be found were we able to compare the structure of all parts of the body; and perhaps future research and more happy discovery may yet bring them to light; at present, however, we are compelled to restrict our comparisons to the wings alone.

First, we have to remark the similarity of the front and hind wings in the ancient types: a similarity which extends to their general form (the extended anal area of the hind wings in modern types being as yet only slightly differentiated); their nearly equal size (a corollary, to a certain extent, of the last); the general course of their neuration (true, in a limited sense only, of modern types); and the complete transparency of the front as well as of the hind wing.

Second, the same number of principal veins is developed in the front and hind wings of ancient Cockroaches; while in the front wings of modern types two or more of the veins are blended, so as to reduce the number of the principal stems below the normal, the hind wing at the same time retaining its original simplicity. These principal veins are six, counting the marginal vein, which here merely thickens the anterior border, as one; to use the terminology of Heer, and starting from the anterior margin, they are the marginal, mediastinal, scapular, externomedian, internomedian, and anal. The general disposition of these veins is as follows:—The mediastinal and scapular veins, with their branches, which are superior (i.e., part from the main vein on the upper or anterior side), terminate upon the anterior margin. The internomedian and anal take the opposite course, and their branches are inferior, or, at least, directed toward the inner margin; while the externomedian, interposed between these two sets, terminates at the tip of the wing, and branches indifferently on either side.

Now these veins are all present in both front and hind wings of palÆozoic Cockroaches, and also in the hind wings of existing species; but in the front wings or tegmina of the latter the number is never complete, the externomedian vein being always amalgamated either with the scapular, or with the internomedian, and the mediastinal frequently blended with the scapular vein.

The hind wings are thus shown to be conservative elements of structure, since they have preserved from the highest antiquity both their transparency and their normal number of veins. They have retained the use to which they were first put, and the changes that have come about, such as the wider expansion of the anal area, have been in fuller development of the same purpose; while the front wings, in virtue of their position in repose, have become more and more protectors of the hind wings, and have gradually lost, in part, if not entirely, their original use. The hind wings of existing Insects, thus protected, have given less play to selective action, and have become to some degree interpreters for us of the more complicated structure, the more modernised anatomy, the more varied organisation of the front wing.

A third distinction between palÆozoic and modern Cockroaches is found in the veinlets of the anal area. These, unlike the branches of the other veins, do not part from the main anal vein at various points along its course, but form a series of semi-independent veinlets, and in palÆozoic Cockroaches take the same general course as the main anal vein, or “anal furrow” (the curved, deeply sunken vein that marks off the anal area from the rest of the front wing, both in ancient and modern Cockroaches), and terminate at sub-equidistant intervals upon the inner margin; while in modern Cockroaches these veins either run sub-parallel to the inner margin and terminate on the descending portion of the anal furrow, or they form a fusiform bundle and terminate in proximity to one another and to the tip of the anal furrow.

These differences, which were mentioned by Germar and Goldenberg, and their universality pointed out in my memoir on PalÆozoic Cockroaches,197 seem to warrant our separating the older forms from the modern as a family group, under the name of PalÆoblattariÆ; this family has been thus characterised:—

Fore wings diaphanous, generally reticulated, and nearly symmetrical on either side of a median line. Externomedian vein completely developed, forking in the outer half of the wing, its branches generally occupying the apical margin; internomedian area broad at base (beyond the anal area), rapidly tapering apically, and filled with oblique mostly parallel veins, having nearly the same direction as the anal veinlets, which, like them, strike the inner margin.

About eighty palÆozoic species have been published up to the present time, and have been grouped in two sub-families and thirteen genera. Besides these, Brongniart has not yet given any hint of how many have been found at Commentry, a French locality which may be expected to increase the number largely, and about twenty undescribed species are known to me from the American Carboniferous rocks.

The two tribes or sub-families differ in the structure of the mediastinal vein; in one type (BlattinariÆ) the branches part from the main stem as in the other veins, at varying distances along its course (see the figure of Etoblattina); in the other (MylacridÆ) they spread like unequal rays of a fan from the very base of the wing (see the figure of Mylacris). What is curious is that the latter type has been found only in the New World, while the former is common to Europe and America. The latter appears to be the more archaic type, since it is probable that the primeval Insect wing was broad at the base, as is the general rule in palÆozoic wings, and had the veins somewhat symmetrically disposed on either side of a middle line; in this case the mediastinal and anal areas would be somewhat similar and more or less triangular in form, and the space they occupied would be most readily filled by radiating veins; such a condition of things, which we find in the MylacridÆ, would naturally precede one in which the mediastinal vein, to strengthen the part of the wing most liable to strain, should, as in the BlattinariÆ, follow the basal curve of the costal margin, and throw its branches off at intervals toward the border, much after the fashion of the mediastinal vein.

This view of the relative antiquity of the two tribes of PalÆoblattariÆ is supported by the fact that while in both of them the internomedian branches show a tendency to repeat the general course of the anal nervules, as in the corresponding veins of the costal region, this tendency is lost in modern types; and among those ancient BlattinariÆ, which are esteemed highest in the series, there is a marked tendency toward a loss of this repetition of the style of branching of the mediastinal and anal offshoots by the scapular and internomedian respectively.

A certain amount of geological evidence may also be claimed in support of this view. A survey of the species of the two groups found up to the present time in America, published and unpublished, shows that all the MylacridÆ are found below the Upper Carboniferous, while more than half the BlattinariÆ are found in or above it. This results largely from a recent and as yet unpublished discovery of BlattinariÆ in the Upper Coal Measures of Ohio and West Virginia, which in their general features are much nearer than previously discovered American Cockroaches to the European BlattinariÆ, the latter of which come generally from Upper Carboniferous beds. The MylacridÆ have therefore been found in America in strata generally regarded as older than those which in Europe have yielded Cockroaches, and this gives a sufficient explanation why no MylacridÆ have yet been found in the Old World. In America one is mostly dealing with absolutely older forms, and they naturally give that continent a more old-fashioned look, when we regard the Carboniferous fauna as a whole. As already stated, a wing from the French Silurian (PalÆoblattina Douvillei Brongn.) has been claimed as a Cockroach, but without good reason, and to see a real old Cockroach one must look to America.

Up to this point we have contrasted the palÆozoic Cockroaches with the existing forms only, and finding such important distinctions between them, we naturally turn with some curiosity to the intermediate mesozoic and tertiary formations.

Now, not only are the mesozoic species as numerous (actually, but not relatively) as the palÆozoic, but a recent discovery of a Triassic fauna of considerable extent, in the elevated parks of Colorado, presents us with a series of intermediate forms between those peculiar to the Coal Measures and those characteristic of the later mesozoic rocks. Excluding, however, for a moment this Triassic fauna, we may say of the later mesozoic species that they are NeoblattariÆ, not PalÆoblattariÆ, though they still show some lingering characteristics of their ancestry. Thus the front wings are in general of a less dense texture than in modern times, but without the perfect diaphaneity of the palÆozoic species; in some the anal veins fall in true palÆoblattarian fashion on the inner margin, while in others which cannot be dissociated generically from them, the anal veins are disposed as in modern types. But in all there is a loss of one of the principal veins, or rather an amalgamation of two or more—a characteristic of more fundamental character. As a general rule, moreover, to which we shall again advert, the mass of the species are of small size, in very striking contrast to the older types.

To return now to the Triassic deposits of Colorado, we recognize here an assemblage of forms of a strictly intermediate character. Here are PalÆoblattariÆ and NeoblattariÆ, side by side. The larger proportion are PalÆoblattariÆ, but all of them are specifically, and most of them generically, distinct from palÆozoic species, and all rank high among BlattinariÆ; still further, the species are all of moderate size, their general average being but little above that of mesozoic Cockroaches, and only a little more than half that of palÆozoic types. The NeoblattariÆ of this Triassic deposit are still smaller, being actually smaller than the average mesozoic Cockroach, and one or two of them, of the genus Neorthroblattina (see figure of N. Lakesii), have marked affinity to one of the genera of PalÆoblattariÆ (Poroblattina) peculiar to the same beds, differing mainly in the union or separation of the mediastinal and scapular veins; while others, as Scutinoblattina, have a Phoraspis-like aspect and density of membrane. This novel assemblage of species bridges over the distinctions between the PalÆoblattariÆ and NeoblattariÆ. We find, first, forms in which the front wings are diaphanous, with distinct mediastinal and scapular veins, and the anal veinlets run to the border of the wing (Spiloblattina, Poroblattina); next, those having a little opacity of the front wings, with blended mediastinal and scapular, and the anal veins as before (some species of Neorthroblattina); then those with still greater opacity, with the same structural features (other species of Neorthroblattina); next, those having a coriaceous or leathery structure, blended mediastinal and scapular, and anal veins falling on the inner margin (some species of Scutinoblattina); and, finally, similarly thickened wings with blended mediastinal and scapular, and anal veins impinging on the anal furrow (other species of Scutinoblattina).

It is not alone, however, by the union of the mediastinal and scapular stems that the reduction of the veins in the wings of later Cockroaches has come about; for in many mesozoic types the externomedian vein is blended with one of its neighbours, while in others not only are the mediastinal and scapular united, but at the same time the externomedian and internomedian.

As regards the other structural distinction between the PalÆoblattariÆ and NeoblattariÆ—the course of the anal nervules—there is much diversity, and very imperfect knowledge, since this very portion of the wing is not infrequently lost, a fracture most readily occurring at the anal furrow. In most of the mesozoic genera, the anal nervules, as far as known, strike the margin; but the larger portion of these show a decided tendency to trend toward the tip of the anal furrow, as in many modern forms. This feature can hardly be considered as firmly established in mesozoic times, and the same genus, as Scutinoblattina, may contain species which differ in this respect.

Fig. 123.—Meso­blat­tina Brodiei Scudd. ×4. Pur­becks, England.