of the maxillæ. This shows clearly how the excavation and the rods and connecting sheets arise by an involution of the cranial wall, with thickenings in special tracts.

The large black ant (Formica pennsylvanica) carries our thoughts still further. It has only one basi-cranial rod (Fig. 4, MS), extending forward above the basi-cranial wall (which is not excavated). This is derivable from the case of the bee, on the supposition that the basal rods and the margins of the basi-cranial involution have approximated medially so as to coalesce. The ant's suspensorium has a medi-suspensorial pair of maxillary rami (MR), as in the bee; but its labial rami are so small as to be nearly obsolete. Its basi-labium (BL) and its medi-labial case (ML) are much as in the bee; but its disti-labial parts are condensed.

The series of gradations thus attained holds out inducements to pursue the subject, and we may perhaps see the beginning of a new line of discovery in this field. Compare, for example, the maxillary adjustments of the cockroach with those of the insects already described. Here again Mr. Huxley is less happy than usual in his anatomical descriptions. He states that the basal piece or cardo of the maxilla of the cockroach is connected with a thin band which runs round the posterior margin of the epicranium and is firmly united with it only on its dorsal side. He does not indeed represent the maxillæ as directly inserted in the back of the skull, but he regards them as attached to a band which is itself attached to the back or dorsal aspect of the skull, and which he is thus compelled to consider a portion of its exoskeleton. This view, if sustained, would clash with the mode of suspension in the bee, where the maxillæ have been found to have endocranial connections with the base or ventral side of the skull.

X19

A careful examination of the conFIG. 6.-Endocranium and maxnections in the cockroach, proves, illary suspensorium of European however, that the cardines of the cockroach (Periplaneta orientalis).

EC, endocranium; CA, cardo. Other maxıllæ are inserted in a ridge which references as in Fig. 5. crosses the basis-cranium in front of

the occipital foramen (though a slender ridge runs from this part

round the occipital foramen, as usually occurs in insects). This transverse ridge is intimately connected with the roots of the mesocephalic system; and may be deemed a condensed variation of the bee's suspensorium.

The Coleoptera have presented the greatest difficulty here, a difficulty which has been long felt by zoologists. With the Coleoptera the basi-cranial region is so unlike that of other insects that a special nomenclature has been devised for it; and the terms mentum, submentum and gula are properly confined to the beetles (the application of these terms to other insects has been, in some measure, guess-work). The base of the head failing us as a guide, we started from the other end, or front. Here it was easy to find in the clypeus of Lachnosterna the points from which the mesocephalic pillars ought to descend; and there the pillars actually are, but they appear as involutions of the wall, and they descend not to the vicinity of the occipital foramen, but further forward to the region of the submentum, and near them the maxillary cardines are inserted. The interpretation of these observations is easy. Mr. Huxley has sought the representatives of the beetle's basi-cranial pieces in the neck of the cockroach; the facts now given appear to say that in other insects (as the bee) they are condensed into the very complex and strong system of ridges which borders the front of the occipital opening. The Coleoptera alone have these parts resolved so as to show the primitive arrangement. The fact that they reach the basi-cranium at the point of insertion of the maxillæ, is in complete harmony with what we have seen in the bee. We observed in the basi-occipital region of the head of Lachnosterna, and still more distinctly in the stag-beetle, an overarching frame, enclosing a nervous canal similar to the sternal canal of the thorax. We may, perhaps, detect traces of this in the very intricate cross-bars in advance of the foramen magnum of the bee; so that here the sternal canal and the roots of the mesocephalic and basi-cranial processes are all crowded together. (Thus it is not correct to say that Coleoptera have no endocranium, although Gegenbaur makes a slip when he cites them as an example of largely developed endocranium.)

Only a few words can be added as to the cranial splachnodemes, or that part of the endocranium which consists of hardening of the pharynx. The mouth is floored by a stiff tongue-like plate

(we may call it lingua, not to be confounded with the ligule already mentioned). The tip of this lingua is deflected downwards, and from its base run backwards two long barbed processes. Over the mouth is a similar but simpler structure, called epipharynx, and to these the borders of the pharynx are attached, and also muscles. If we force open the mouth (by pulling down the maxilla), we find the open mouth overarched by epipharynx (connected with the labrum), floored inwardly by the lingua (or inner tongue, formed by the floor of the pharynx), enclosed at the sides by the long tendons of the lingua which are stretched up like faucial door posts. All these hard parts keep open the soft membrane of the pharynx, just as the iron frame of a dredge keeps open the netting attached to it. In the upper part of the cranial cavity are racemose glands which send down a pair of ducts to the inner tongue. The great salivary apparatus of the thorax sends forward its ducts which unite and penetrate through the basi and medi-labium to the ligule or long outertongue.1

It would be premature, at the present stage of our knowledge, to theorize upon these facts. They indicate a fundamental unity of structure of the heads of all insects; but how far and in what directions it is varied, and what is its relation to other parts of the body, are questions needing further research.

:0:

IN

MYA ARENARIA IN SAN FRANCISCO BAY.

BY ROBERT E. C. STEARNS.

́N November, 1874, Dr. W. Newcomb, who at that time resided in California, described in the Proceedings of the California Academy of Sciences, a species of Mya which had been given to him by the well-known collector, Henry Hemphill, who detected several specimens of the form on the shore of Alameda county, on the eastern side of San Francisco bay.

The specimens were about two-thirds of the usual average size

1 Siebold discovered a triple salivary system in the bee; but the text books are still sadly at variance with each other and with the facts, in their treatment of this part of the subject. Some place the bee's salivary glands in the head, some in the thorax, and some say they are sometimes in one part and sometimes in the other!

of Mya arenaria, rather fragile in substance and delicate in sculpture.

As Dr. Newcomb considered it a new species, he described it as above and gave it the name of M. hemphilli, remarking that "the only species with which it can be confounded is the Mya præcisa of Gould, which Dr. Carpenter considers as identical with M. truncata of the North Atlantic. A specimen of M. arenaria, from Puget sound, in my collection, is quite distinct from this species, and, like many of the circumpolar species, is common to the North Pacific and North Atlantic. It is quite distinct from the fossil M. montereyi Conrad, as I am informed by Dr. Cooper, who kindly made for me the comparison of this shell with Conrad's figure and description."

Since 1874, the date of the description, the Mya has become abundant, and is found for miles alongside the easterly shore of the bay, and is now the leading clam in the markets of San Francisco and Oakland, superseding to a great extent the previous "clams," Macerna nasuta and Tapes (or more properly Cuneus) staminea Conrad, in its varieties, especially diversa Sby., and the now dominant clam of the fish-stalls, is found exhibiting all of the characteristics of Mya arenaria, and is universally conceded to be the same as the Atlantic species.

My friend, Dr. Newcomb, as quoted above, it will be seen, regarded his Puget sound Mya as M. arenaria; which is the region from which Gould's form, Mya præcisa, was brought, and if related to another form, is more likely to be a variety of the circumpolar truncata than to be arenaria.

None of the more recent and reliable collectors referred to in Carpenter's Supplementary Report (1863) to the British Association, neither any collector since this date, to my knowledge, has verified the occurrence of M. arenaria at any point north of San Francisco bay on the west coast of America. Gould was certainly familiar with a form so common on the New England coast; and though perhaps in this day and generation we hold rather broader views as to what constitutes a species than some of the old mas

The only bivalve along the coast or in the bay of San Francisco which might be mistaken for or identified with Mya arenaria is Schizotharus nuttalli Conrad, which when mature is two or three times as large as the largest specimens of Mya. Dwarfed forms of the large species are found at low tide on the flats connected with Goat island on the east. In this species the siphons are enclosed in an external sheath, the same as in Mya, making what the unsophisticated call a “long neck."

ters, yet he certainly would have noticed it, had the shells before. him been closely like the more common Linnean species.

Middendorff credits it to Sitka, etc., but this is not supported by any of the numerous subsequent authorities.

As to its presence on the Asiatic side of the North Pacific, Middendorff credits it to "Kamchatka" and the "Ochotsk sea." Jay, in (Vol. 11) Perry's Japan Expedition, describes it even as "M. japonica," and credits it to "Volcano bay, Island of Yeddo," remarking that it "is very similar to M. arenaria." Arthur Adams, who collected in Japan, pronounced it identical. It has also been detected in Hakodadi bay, and Professor Morse says "the typical northern form (M. arenaria) lives in the Gulf of Yeddo to-day, and its shells are found in the mounds of Omori.”3

It must be admitted that the species is found in these Asiatic stations upon the testimony here adduced, but as to its presence on the coast of Western America at any point north of or anywhere outside of San Francisco bay, the fact that neither Nuttall, Jewett, Kennerly, Lord, Swan, Cooper, Harford, Dunn, Hemphill, Hepburn, Fischer, Dall, Newcomb (in the field), and many others, as well as myself, have never detected a specimen prior to the date of my friend Newcomb's description (or since, so far as I can learn), ought to be sufficient evidence on this point.

From whence, then, came the seed which has produced the abundance of this species which has spread and is now spreading rapidly along the shores of San Francisco bay?

Examine the ancient shell heaps and mounds found hereabout, and one may find the thin broken valves of the Macomas, but not a fragment of the shell of Mya. One may find the shells of the native Haliotis and Olivella and the beads and money or ornaments made from them; the bones of the common California deer, of the whales, and perhaps other animals, all of which are still to be found in the neighborhood or not many miles away, but not a piece of Mya. The ancient clam-diggers, whose kitchenmiddens are met with in many places on the Alameda and other shores of the bay, whose skeletons and implements are sometimes exhumed or discovered, had "passed over to the majority" centuries before the advent of Mya arenaria in California waters. To proceed to the question-was the seed of this mollusk intro1 B. A. Report, 1856, p. 219.

2 Id. Rep., 1863, p. 588.

3 AMER. NATURALIST, Sept., 1880, p. 657-8.