52 



KR. BONNEVIE. 



[REP. OF THE "MICHAEL SARS" NORTH 



the age of the individual, and their shape, constant within 

 each species and sometimes within each genus, is deter- 

 mined by characters so relative (more or less broad basal 

 plates, larger or smaller hooks, etc.) that a system based 

 upon such characters might easily lead to confusion. 



Besides the median tooth of the radula, the jaw proves 

 to be a "good" family character. In this particular also 

 the two families Pneumodermatidae and Notobranchaeidae 

 show absolutely different types, the jaw in the former 

 being represented by a conical or cylindrical papilla, while 

 in the latter it is found in the form of a row of small 

 spines placed directly on the ventral wall of the buccal 

 cavity. In those cases where the median tooth of the 

 radula is too much reduced to manifest the family type 

 (as in the genus Microdonta), the jaw may be typical 

 enough to prove the relationship. 



The want of a jaw is characteristic of the family 

 Clionidae, while the type of a jaw in the Clionopsidae 

 has not yet been determined. 



The systematic importance of the radula and jaw, 

 is so well known in other groups of molluscs that their 

 neglect in the gymnosomatous pteropods is probably due 

 to the difficulty of the determining their form in con- 

 tracted animals. 



Besides these old mollusc-organs there are others, 

 characteristic of the gymnosomatous pteropods, more easily 

 recognizable and therefore playing a greater part in 

 systematic works, such as the hook-sacs, the buccal appen- 

 dages and the gills. 



In table IV the systematic value of these organs is 

 tested by comparing their uniformity, within smaller or 

 larger groups of species, with that of the jaw and 

 radula. 



Considering first the gills, which are considered so 

 important especially in the system of Pelseneer, we shall 

 find them to be very inconsistent organs, and therefore 

 not fit to serve as the basis of a natural system. The 

 lateral gill might perhaps be of some importance in so 

 far as it has not been shown to exist outside of the family 

 Pneumodermatidae, but even within this family it may be 

 absent (Schizobrachium) , or when present it may vary so 

 much in size and shape that no more than a specific 

 value can be attached to it. The same is, to a still greater 

 extent, true of the posterior gill, which has been found 

 in all the families except the Clionidae, but in each of 

 these families it may be present or absent, and when 

 present it may consist of radial crests varying in number, 

 of a membranous ring, or of both structures together. 



More constant organs than the gills are met with in 

 the hook-sacs, the type of which may be constant within 

 whole families or at least within large groups of species, 

 but the same type is met with in different families, and 



no more than a generic value can therefore be attached 

 to this character. 



Besides the organs enumerated, which may practically 

 be considered characteristic of the whole group of gymno- 

 somatous pteropods, there are other organs characteristic 

 of one family only, such as the acetabuliferous appendages 

 of the Pneumodermatidae, the long thread-like proboscis 

 of the Clionopsidae, the buccal cones of the Clionidae, 

 and the complex proboscis of some Notobranchaeidae. 



A natural system of gymnosomatous pteropods should 

 accordingly be based in the first place on the old mollusc- 

 organs, the radula and jaw, in the second place on the 

 new acquirements of the order or of certain families, viz. 

 the hook-sacs and the special characters of the buccal 

 mass, which may be important as having generic value, 

 and finally on the adaptive characters of foot and gills, 

 which are of great value in the description of species. 



In the following descriptions I shall give the charac- 

 teristics of the organs mentioned, in so far as they play 

 a part in the diagnoses of families or genera. 



Pneumodermatidae. 



Radula: Median tooth with 3 denticles; in species 

 where it is wanting in the adults, it may be present in 

 the young stages (textfig. 39), 



Jaw: A papilla, conical or cylindrical, wholly or 

 partly covered with spines. 



Hook-sacs: 4 different types corresponding to the 

 genera. 



Acetabuliferous appendages present. 



Laterallgill generally present. 



Pneumodermopsis Bronn. 



Radula and jaw of the family type, 



Median tooth of the radula well developed. 



Hook-sacs shallow with short hooks. 



Acetabuliferous appendages consisting of a 

 pair of lateral symmetrically developed arms and a median 

 part varying in appearance. 



Pneumodermopsis macrochira Meisenheimer. 

 PI. V, fig. 35-44. 



The description of this species given by Meisenheimer 

 (1905) is based upon two individuals from the "Valdivia" 

 expedition, both contracted, so that several characters had 

 to be left undescribed. 



During the "Michael Sars" expedition 24 individuals 

 were preserved of a species which in essential point agrees 

 so well with the short description of Meisenheimer, that 

 I do not hesitate to refer it to that species, although there 

 may be some minor differences. 



