CUMACEA FEOM THE COPENHAGEN MUSEUM. 343 



From the point of view of the general taxonomy and phylogeny of the Crustacea the 

 results of recent additions to our kno\A'ledge of the Cumacea are disappointing. Not one 

 of the new-genera which have been described perceptibly extends the group towards 

 any of the adjacent Orders of the Malacostraca. The separation of the branchial 

 orifices in Zygosiphon and Schizotrema might perhaps be thought to take us a little 

 way back in the direction of the Mysidacea, but it is more likely to be a secondary 

 modification than a primitive character ; and although Ceratocuma in its free telson 

 and full series of pleopods unites two characters, presumably primitive, not elsewhere 

 found in combination, it is in all other respects a specialized type. The Cumacea still 

 remain a sharply circumscribed group, and although it is as certain as anything of the 

 sort can well be that they have been derived from a Mysidacean-like ancestral form 

 and that their line of descent travelled for some little way along with that of the 

 Tanaidacea and Isopoda (but not, probably, with that of the Amphipoda), none of the 

 intermediate links appear to have survived. 



It was to be expected that the classification of the Cumacea, established by Sars 

 almost entirely from a study of northern species, would require modification as a result 

 of the discovery of so many new species from other seas. As usually happens in such 

 cases, the diagnoses of existing families have to be greatly modified if they are to admit 

 the new forms, and the limits between the families become harder and harder to 

 recognize. I have elsewhere proposed the union of the Campylaspid^ with the 

 Nannastacidas, and have commented on the increasing difficulty of keeping the 

 Vauntompsoniidee apart from the Bodotriidse and the Diastylidte from the Lampropidse. 

 The new genus Colurostylis described below similarly helps to bridge the gap between 

 the Diastylidse and Pseudocumidse. In many cases characters hitherto relied on for 

 the separation of families — such as the number of thoracic exopods or of pleopods — 

 have proved to be of not more than generic value, and in the case of some of the species 

 of 'Sannastacus described below I have been led to suspect that the number of thoracic 

 exopods may not be even a specific character. Whether it will be possible to group 

 the genera of the order in a smaller number of natural and sharply defined families is 

 a problem which must be left for the future. It is at least as likely that it will be 

 necessary, as it has been in the case of the Gammaridean Amphipods, to establish a 

 much larger number of families and to be content to define these by more trivial 

 characters. At the present moment, however, any attempt at a re-classification could 

 only be of the most provisional nature, and I have therefore preserved, where possible, 

 the existing families, with such extensions as are necessary to admit the new forms. 



In the descriptions which follow, the measurement of " total length " includes the 

 telson (when distinct) but not the uropods. As it is often impossible to straighten 

 out the abdomen without the risk of damage to the specimen, it is difficult to obtain 

 accurate measurements, and the figures given must be regarded only as approximate. 



3a2 



