1 82 Southern Cross. 



NOTOTHENIA, Eichardsoii. 



The identification of the fishes of this Peri- Antarctic genus is 

 beset with great difficulties, in the absence of a general account of 

 them, and this deficiency has resulted in the description of a good 

 number of purely nominal species. An attempt to reduce the 

 number of species described from the Fuegian district has recently 

 been made by Professor Smitt,^ but a careful examination of the 

 large material in the British Museum has failed to convince me of 

 the soundness of Jiis conclusions. In no species do I find the number 

 of rays in the anterior dorsal varying as much as from 4 to 7, as 

 Professor Smitt believes, nor does the size of the eye prove to be 

 an unreliable character, provided the comparison be made of similar- 

 sized specimens ; and the width of the interorbital region, compared 

 to the- length of the head, is still more important as not varying with 

 age. Autopsy of a number of individuals among the material at 

 hand has satisfied me that sexual dimorphism does not account for 

 the differences in the proportions of the parts of the head which has 

 been ascribed to it. In his endeavour to settle questions of species 

 by means of elaborate tables of measurements, the Professor appears 

 to have lost sight of many really important specific characters, and I 

 consider his contribution as a misleading guide to the study of this 

 difficult group. My conclusions are embodied in the following 

 synopsis of the species, concerning which I wish to point out that I 

 have myself counted, with the greatest care, the fin-rays and scales in 

 nearly all the specimens at my command. Except for the definition 

 of N. canina and N. filholi, which are only known to me from the 

 descriptions, all the characters are taken from specimens in the British 

 Museum. 



The length of the head is measured to the extremity of the 

 opercular bone. The scales in a longitudinal series are counted above 

 the lateral line, from the origin of the latter to the end of the 

 muscular part of the tail ; those in the transverse series are counted 

 from the middle of the anterior dorsal to the lateral line, and from 

 the latter to the mid-ventral line, some distance in front of the vent. 

 The lateral line, in these fishes, consists partly of tubules and partly 

 of more or less distinct impressions or pits ; only the tubular scales 

 are counted. For the purpose of uniformity, the scales have been 

 almost invariably counted on the left side of the specimens. 



1 Bih. Vet. Ak. Ilandl., xxiii., iv., 1897, No. 3. 



