678 MATNARD M. METCALF, 



in the case of an all)^ such as Octacnemus, which was attached, or 

 at least not locomotor, but it is difficult to see why the stigmata in 

 the walls of the sac should become closed up, unless perhaps nu- 

 trition and aeration were performed sufficiently by the water gaining 

 access to the large cavities of the body through the branchial and 

 atrial apertures, without their being any definite current." 



The only special point here mentioned in which Octacnemus 

 agrees with Salpa in distinction from the other Tunicates is in the 

 aggregation of the viscera into a compact mass. On the other hand 

 Octacnemus differs from Salpa iu the other features which are 

 especially characteristic of the Salpidae ; namely the absence of lateral 

 walls to the pharynx, the presence of an eye, the nature of the 

 neural gland, and the peculiar relations of the reproductive organs 

 in the solitary individuals and those that reproduce by budding. If 

 the features of resemblance between Salpa and Octacnemus were of 

 sufficient importance, Herdman's explanation of the differences in 

 the pharynx might perhaps be adequate. As it is, however, it hardly 

 seems convincing. Let us refer to the other points. 



Octacnemus has no eye. All the Salpidae have a highly de- 

 veloped eye, or often several eyes. This difference might also be 

 accounted for, as Herdman accounts for the differences in the 

 pharynx, by the locomotor habit of Salpa, and the attached condition 

 of Octacnemus. 



The neural gland of Salpa is unique, consisting of two lateral 

 chambers each connected with the pharynx by a tubular duct. In 

 Octacnemus we find a neural gland which, in its anatomy and histo- 

 logy, resembles the neural gland of the Clavelinidae and the 

 compound Ascidians, exclusive of the Botryllidae. 



In Salpa we have two kinds of individuals, one the solitary 

 form, having only a mass of immature germ cells; the other, the 

 chain form which is produced by budding from the solitary form, 

 and which receives its germ cells from its parent and matures them, 

 serving as a nurse. The Challenger specimens of Octacnemus hithyus 

 showed no indication of any stolon. They may have been individuals 

 that had been formed by budding and had separated from the parent. 

 As, however, there is nothing to indicate such an interpretation we 

 must accept Octacnemus hithyus, either as a solitary species, or the 

 Challenger specimens as non-budding individuals of a species which 

 sometimes buds. 0. patagoniensis on the other hand reproduces by 

 budding. But both 0. hithyus and 0. patagoniensis have well de- 



