3. Nymphaea flava Leitner.-The plate of Audubon's great work which represents the American swan likewise represents the flower of a yellow Nymphea, or true water-lily, under the above name. The foliage which accompanies it may be that of a Nuphar, but the flower is that of a Nymphoa. Leitner was a German botanist who explored southern Florida, and died or disappeared there, -if we rightly remember, was thought to have been killed by Indians. He doubtless met Audubon and gave him the name which he published on his plate. The species has properly been left unnoticed so long as the whole evidence of its existence rested upon Audubon's figure of a flower accompanied as it is with Nuphar foliage. But of late years we have heard of a yellow water-lily in Florida. In 1874, Dr. Edward Palmer sent us a specimen with foliage and flowers collected in Indian River, and certified to the yellow color. It has now been detected by Mrs. Treat, on the St. John's River, and living plants communicated to us, from which we may expect to see fresh blossoms. The growth is very different from that of N. odorata, the rhizoma being shorter, and thickly beset with salient blunt tubercles; and the plant propagates freely by stolons.

A. G.

4. Note on some of the Starfishes of the New England Coast; by A. E. VERRILL.-In the Archives de Zoologie Experiméntale et Générale, vol. iv, Nos. 2 and 3, 1875, M. Edmond Perrier has published a very useful and important paper entitled "Révision de la Collection de Stellerides du Muséum d'Histoire Naturelle de Paris," in which he has redescribed many of the types of Lamarck, J. E. Gray, Müller and Troschel, and others, as well as many new species, and has also added many remarks on various genera and species, as well as upon their classification, etc. At the present time I do not propose to discuss this memoir, as a whole, but wish to call attention to some errors into which the author has fallen concerning our common New England species, owing chiefly, doubtless, to his not having a sufficient number of well preserved specimens to form any clear ideas of their true specific characters and great variability.

Every naturalist who has occasion to collect and study any considerable number of living specimens of any of the larger species of Asterias, especially if from different localities or varying stations, must be deeply impressed by their extreme varia bility, not only in size and color, but in the form and relative length of the rays, character of the dorsal spines, number of pedicellariæ, etc. Moreover, if he has had occasion to preserve large numbers of specimens, both in alcohol and by drying, he must have observed the very different forms and appearances that specimens, quite similar when living, will assume, whether owing to the various states of contraction in which they die, or to the mode in which they are afterwards preserved. Thus similar living specimens may be killed and preserved so that one will have slender tapering rays; another, rays smaller in the middle and constricted at base; another, rays swollen at base and pointed at

tip; some will die with swollen disk; others with contracted disk; some will have the rays collapsed and flattened; others will have them round and plump, or angular; some will have the spines erect; others, more imperfectly preserved, will have them flattened. down and more or less detached. Moreover, the plates in some will be so closely drawn together by the contraction of the muscles of the skin as to give them a rigid character, while others, perfectly identical. if they die in a relaxed or inflated condition will have the plates separated by the looser integuments so as to give them an openly reticulated appearance, with wider naked spaces between the plates. Hence all such characters should be used with great caution.

It is, therefore, evident that any naturalist who would correctly limit the species in this group should at least have a very large number of specimens preserved, as well as possible, in various ways, and still better, when possible, he should collect large numbers of the living specimens and after studying them in life and making notes upon them he should preserve, and afterwards compare them with his notes. In this difficult group there are probably no species more variable and perplexing than those forms allied to Asterias rubens of Europe, and A. vulgaris and A. Forbesii, the common New England species. And yet in this very group M. Perrier attempts to decide the specific characters of our species, and to correct their synonymy after an examination of very few (sometimes only one), and often very badly preserved dry specimens (A. pallidus). And in doing this he relies on characters that are notoriously variable, and even upon those accidental features due to modes of preservation, as stated above.

As M. Perrier particularly refers (pp. 354-7) to my own views in regard to our native species, as expressed in several former papers,* and seems to think it strange that my conclusions in 1873 differed slightly from those held in 1866, I may be pardoned for stating that during the ten years that have elapsed since my first paper on the subject was published, these starfishes have been collected, studied, and preserved by me in very great numbers, and from hundreds of localities, during the various dredging expeditions that I have undertaken along our coast, some of which have been noticed in former volumes of this Journal. Therefore, having carefully examined many hundreds of specimens, in all conditions, and having taken ten years to consider the matter and to discuss it with others, I feel perfectly confident that M. Perrier has made at least five American " species out of specimens that actually represent but two. These errors would have been more excusable had they not been made subsequently to my revision of the synonymy, for he might have supposed that my materials were far more ample than his own. The facts are as follows: Upon the coast of New England there are, as yet known,

* Proceedings Boston Society of Natural History, vol. x, p. 333, 1866; Report on the Invertebrata of Southern New England, Report of U. S. Commission of Fish and Fisheries, Part I, 1873 (published March, 1874).

but three species of Asterias, belonging to the littoral zone and shallow waters, though two or three other smaller species (Leptasterias) occur in deep water.* One of the shore species, A. littoralis (Stimpson), is a small species, rarely six inches in diameter, (belonging to the group, Leptasterias) found in the Bay of Fundy and northward to the Gulf of St. Lawrence, both between tides and in deep water, and although variable in form, size, color, etc., it has not yet led to much confusion. The two others are large and very common species-one southern and the other northern-but with their respective areas overlapping on the New England coast. One of these, A. Forbesii (Desor), extends from the Gulf of Mexico, to Casco Bay, Maine, and is the most common species on the southern coast of New England. The other, A. vulgaris (Stimp.), extends from Labrador (and probably farther north) to Long Island Sound, where it becomes rare; it is the most common species north of Cape Cod, and is very abundant on the coast of Maine and northward. Both species vary extremely in size, form, and color, even when living, and still more so when preserved by drying. The color of the faded dry specimen, which M. Perrier mentions as distinguishing "A. Fabricii" from "A. berylinus," is a perfectly worthless character, yet when living our two species can usually (but not always) be distinguished by the colors, for A. Forbesii (berylinus) is generally greenish, varying to orange and brown, with a bright orange madreporic plate; while A. vulgaris is usually reddish, purplish, or violet, varying to yellow and brown, but with a pale buff or cream-colored madreporic plate. Moreover, the colors and forms of each vary according to the sex, and the form varies according to the season, state of the ovaries, age, dilation with water, etc. The forms of the major pedicellariæ, and of the adambulacral (or interambulacral) spines give the most available characters for distinguishing the two species under all circumstances, though the firmer skeleton of A. Forbesii is also an important and characteristic feature.

In A. Forbesii the major pedicellaria are short, ovate, bluntpointed, hardly longer than broad; the adambulacral spines are stout, obtuse, and in most cases many are more or less flattened, and grooved externally, at tip.

*It is proper that I should state that I have become satisfied that the species described by me as A. Stimpsoni, in 1866, was not well founded. The study of a far more extensive series of specimens has shown that the specimens thus named were somewhat peculiar small specimens of A. vulgaris (Stimp.), with which some young specimens of A. littoralis were also confounded, so that the characters given largely appertain to the young of A. vulgaris. It is probable that most of the specimens formerly distributed as A. Stimpsoni, were young of A. vulgaris, and such may have been those that M. Perrier says he has examined, though he considers it a good species.

+ Dr. Stimpson, in 1862, (Proc. Boston Soc. Nat. Hist., vol. viii, p. 262, note) called special attention to the two peculiar forms of pedicellariæ, characteristic of this family, and to their importance in distinguishing the species. He termed the larger sessile ones, whether scattered over the surface or attached to the ventral spines, "major pedicellaria," and the small pedunculate ones, borne in clusters on the spines, "minor pedicellaria." In 1864-70 M. Perrier called the former, "pédipellaires droits," and the latter "pédicellaires croisés."

In A. vulgaris the major pedicellaria are lanceolate, sharppointed, much longer than broad; the adambulacral spines are longer, more pointed, and seldom flattened.

The dorsal spines are variable in form and number in both species, but are usually more acute in the latter, though blunt and even clavate spines often occur on both species. The number of minor pedicellariæ on the spines, and of major ones on the back and also on the adambulacral spines, is extremely variable in both species. Yet these are the characters mainly relied upon by M. Perrier for distinguishing his supposed species. Three of the "species" recognized by him are evidently mere forms of A. vulgaris. These are A. Fabricii (Agassiz, MSS.), based on one dry specimen from Laborador; A. pallidus (Agassiz, MSS.) based mainly on six small dry specimens, sent, like the preceding, from the Museum of Comparative Zoology in 1864, and in a very bad state of preservation, the spines being mostly detached by partial decomposition during the drying of the specimen (which has misled M. Perrier, who imagined that they had been moveably articulated); and two large specimens that he refers to A. vulgaris, one from Beverly, Mass., and one in the British Museum. The latter was probably sent by Dr. Stimpson, who commonly used labels marked "Exploration of the East Coast of the United States" (not "west" coast, as M. Perrier gives the label), for his New England collections. I had given "A. Fabricii," after examining original specimens, as a synonym of A. vulgaris in my paper of 1866; and gave " A. pallidus" as an undoubted synonym in my Report of 1873-4. All the characters given by M. Perrier as distinctive are variable and partly accidental features that can be found, with all intermediate states, in any considerable collection of this species from a single locality. These manuscript names were given by Prof. L. Agassiz before he had made a very thorough study of the genus, but in a conversation, in 1871, while we were dredging in company in Vineyard Sound and obtaining both species in abundance, he fully agreed with me that there were only two large species of this group known on our coast, and he also positively identified the numerous good specimens of A. vulgaris with his A. pallidus; and likewise the A. Forbesii (or arenicola) with his A. berylinus. Dr. Wm. Stimpson in a conversation with me not long before his death, also agreed with these decisions. Had M. Perrier examined a good series of specimens he also would surely have found it impossible to have made the useless distinctions that he now proposes, based on such very insufficient material. In my Report of 1873-4 I stated that A. Forbesii and A. arenicola are "probably identical," the differences noticed (mainly in form and color)" being, perhaps, chiefly sexual," but not desiring to make premature changes, I left them under the two names, only because I had then no time to determine whether the differences are sexual, or properly varietal, or due to local or individual variations. Subsequent studies have satisfied me that the differences are mainly individual or casual and very incon

stant, so that there is no longer any reason for distinguishing the two even as varieties, yet M. Perrier not only keeps them distinct, as two species, but names another slight variation as a distinct variety of A. arenicola.

There is now little doubt in my mind that A. vulgaris will prove to be identical with A. violaceus of Northern Europe, and that the latter may be, as many believe, a mere variety (or sexual form) of A. rubens. But M. Perrier considers these distinct species, though with some doubt as to A. violaceus.

That, in other cases, he has admitted, as valid species of Asterias, forms that are scarcely varietal is very probable, judging from his descriptions alone, for the distinctive characters that he gives are frequently those that are most apt to be variable. He has described a single dry Labrador specimen of A. polaris (from Dr. Packard's collection) as a new species under the name of A. borealis. But among a large number of fresh specimens observed by me at Anticosti Island, there were various forms intermediate between his specimen and what he regards as the typical A. polaris from Greenland. Moreover, the several Labrador specimens that I have examined, collected by Dr. Packard at the same time with the one now described as A. borealis, show great variations in the form of the spines, length of armis, and number of pedicellariæ,-characters that M. Perrier regards as distinctive in this case. Therefore there is good reason to believe that his A. borealis is only a form of A. polaris, to which American zoologists have hitherto referred it.

Throughout the paper there are numerous typographical errors, many of them due to imperfect proof-reading, but others more important are due to careless references to the papers of other writers, especially those in English. "Contributions to the Zoology of Yale College" is scarcely a legitimate substitute for the "Contributions to Zoology from the Museum of Yale College," published in this Journal. The locality, Eastport, Maine, is once given as "East Port (Massachusetts,)" and once as "East Port, (Canada.)"

In one respect the nomenclature adopted, in some cases, by M. Perrier is very objectionable, for he attempts to restore some of the ante-Linnæan "names" of species used in 1733 by Linck, who was not, in any sense, a binomial writer, and whose polynomial (or accidentally binomial) phrases can have no claim to priority, as specific names, under the binomial system.

5. Hæckel's theory (Alloogenesis) of the genetic connection between the Geryonida and Eginida.-In the Proceedings of the Elliot Society for 1857, McCrady gave a very interesting account of the commensalism of the young brood of a Cunina and of Turritopsis. No notice was taken of this remarkable mode of development, McCrady's observations having been discredited by the later publication (1865) of a magnificently illustrated memoir on the "Rüssel-quallen" by Hæckel. The startling hypothesis of the genetic connection between the Geryonidæ and Æginidæ contained