l^ lAdh: si\(i(;si!i;i;(; 



AlttT till' cliininatioii ol tln-sc ilivt-r^ii'iil I'lciiu'iits the i;cnii.s ('i//in(lin(i (sciisii 

 I J. \\ . MCl.l.KlU) iniiy fortainly bo siiitl to liavc Ix'comc (■(nisidcraljK' luoic uniform. 

 A furtlicr division of it appears, however, to lie [larticularh drsiialilr, Tlir in(iitn|)lcicness 

 and uncertainty of tlie diagnoses and Hfjures of t hr ma jnrit \- nl i he icmaminu sprcics aic however, 

 so great that any attempt at the present moment to cair\ out a natural ai ran^cinrnt and division 

 of all these forms must he eonsidereti premature and inconvctdcnt. 



A fairly large number of species belonging to this genus were found in tiie material that 

 formed the basis of the present treatise. Because of this it seemed to me necessary, in spite 

 of the difficulty of the task, to attempt even now tocarrv out a partial di\isi((ii of the genus in 

 question. In doing this I have, of course, chielly taken into account tlic forms tliat I myself 

 have bad occasion Ui investigate closely, i have tried, however, as far as possible, to show the 

 relations of these forms to species that have l)een previously dealt with in the literature. 



As a result of this attempt I have established four new sub-genera, as shown below; tiiesc 

 sub-genera are: 



Doloria 

 Vnrgulu 



Macrocypridina 

 Sipho nostra. 



The sub-genus Vanjulu is based on three species investigated by me but ahead)- described 

 previousl\% namely Cypridina norvegica W. Baihd, C. antarctica G. W. MOLLER and C. megalops 

 G. 0. SAIiS. Besides these three species a large number of the forms included by G. W. MOLLER 

 under the name of Cypridina are probably to be referred to tliis sub-genus. That under these 

 circumstances the name Cypridina has not been retained for these species, but that in the present 

 work this name has been made to replace the entirely rejected generic name Pyrocypris, will 

 be found explained below; see note under the sub-genus Cypridina. — Macrocypridina 

 comprises only one species, Cypridina castanea established by G. >S. Bradv, 1897. — The 

 two remaining sub-genera, Doloria and Sipkonostra, are based on species previously unknown 

 to science. WTiether any previously described species belongs to the sub-genus Doloria, I cannot 

 say. To the sub-genus Siphonostra possibly belongs, besides C. (S.) spinifera described below, 

 Cypridina nobilis P. T. Clex e. With regard to the relationship of the last-mentioned sub-genus 

 to Cypridina hirsuta G. W. MCller and the extremely incompletely known genus Heterodesmus 

 established by G. S. Bhady see below, note under the sub-genus Siphonostra. — With regard 

 to the characters by which these sub-genera are distinguished I merely refer here to the diagnoses 

 of these groups given below. 



It is true, on the one hand, that all these four sub-genera are undoubtedly fairly closely 

 related, but on the other hand they show such great differences from each other that it seems 

 to me quite correct to distinguish them. I have had some doubt as to whether it is most 

 convenient to denote these groups as genera or sub-genera. The relatively close relationship 

 that exists between these groups in comparison with most of the previously established genera 

 of this sub-family has induced me to put forward these new units as sub-genera. The question 

 is, however, not a very important one 



