28 BULLETIN OF THE BUREAU OF FISHERIES. 



different condition of deep-sea life, the burden of proof rests upon those who make 

 the claim. Whether available or not for animals with functional visual organs, 

 however, it fails to touch the problem among creatures which, like many of the 

 medusae, are wholly devoid of such organs; but interesting as might be further dis- 

 cussion, the. present is not the most appropriate place for the treatment of this and 

 related problems, nor is their morphological aspect likely to afford more than mere 

 plausibility. As suggested in a previous paragraph, it seems to me to be a 

 physiological rather than a morphological question. This phase lias been discussed 

 by me elsewhere (cf. Science, Jan. -2-2, 1904), and the briefest summary is all that 

 can lie attempted here. 



It was long ago pointed out by Darwin himself that color among lower forms 

 could not have any value as a factor in natural selection; it was considered as an 

 expression of the "chemical nature or minute structure of their tissues," and Wallace 

 regarded color as "a normal product of organization v among such creatures. It has 

 more recently been determined that among annelids, mollusks, and. indeed, among 

 certain insects, coloration often results from the purely physiological processes of 

 excretion, and I have shown that the same is probably true concerning the coloration 

 of medusae and other ccelenterates. Morgan and Stevens have demonstrated that 

 among hydroids during regenerative processes pigmentary matters, which were first 

 thought to have an important function, were in reality waste products in process of 

 elimination. The marked changes of coloration often observed in these organisms 

 under artificial environment have been shown to be associated with changed conditions 

 of nutritive metabolism in some instances, while in others light has been found to be 

 an important factor. 



Similar observations and conclusions are not lacking concerning coloration among 

 echinoderms, Crustacea, and many other groups. Without seeking to discredit the 

 value of color as a factor in adaptation among some of the highly specialized groups, 

 I can not repress the conviction that its importance in this respect has been greatly 

 overestimated. 



SYSTEMATIC SYNOPSIS. 



The medusae comprise three fairly well differentiated classes, known as Hydromedusae, Scypho- 

 medusae, and Ctenophora, ami distinguishable by the following diagnostic characters. 



I. HYDROMEDUSjE. — Definite velum, gonad:- usually ectodermal and developed upon the 

 walls of the manubrium or beneath the radial canals; sensory organs in the form of ocelli or otocysts, 

 never modi lied tentacles. Polyp stage usually present, but devoid of gastric filaments or ridges, and 

 producing medusae or sporosaes by a process of budding. 



II. SCYPHt iMKDI'S.E. —Devoid of true velum; gonads always entodermal, and borne within the 

 gastric pouches; sensory organs when present are usually entodermal and are perhaps modified 

 tentacles. Polyp stage when present possessed of distinct gastric ridges and often filaments, and 

 multiplying or metamorphosing by transverse tissiou to form the larval medusa. 



III. CTENOPHORA. — Devoid of nematocysts, locomotion by means of vibratile plates, which are 

 arranged in eight meridional rows. Tentacles when present only two in number, adradially placed 

 on opposite sides of the body. 



THE HYDROMEDUSjE. 



The Hydromedusae comprise some live fairly distinct orders: 



I. An ruoMKiu's.K. — A more or less hemispherical umbrella, sometimes ovoid or conical; velum 

 usually well developed, affording an effectual swimming organ, because of which these medusae are 



