54 HIALMAR RENDAHL 



head; interorbital space 0,9 length of eye. Fin formula: D. 5/61, A. 57, P. 

 24. V. 9. 



Steindachner first identified the Lotella from Juan Fernandez as L. 

 rhacinus Forst.; subsequently he »nach der Zahl der Ventralstrahlen» referred 

 it to L, physis Schleg., previously known from Japan. I think, however, that 

 it is distinct from both these species. Unfortunately I have no material of the 

 two mentioned species for comparison and must base my opinion only upon 

 the rather short descriptions in the literature. 



From L. physis the proposed new species may be distinguished by its 

 broader interorbital region. The width of the latter in L. physis is less than 

 the diameter of the eye. Steindachner and Doderlein, who redescribed 

 the species (Denk. Akad. Wiss. Wien. LlII, 1887), based their statements on 

 seven samples, the length of which they do not mention. One of the speci- 

 mens is, however, stated to be 21 ^/a cm. 'Wx^ ^iXO'^oxWon interorbital space : eye 

 is of course a characteristic, that varies with the size of the specimen. In 

 my specimen, which is not more than 15 ^/2 cm., the width of the interorbital 

 space exceeds, however, the diameter of the eye, whereas it in the smaller 

 sample, 11 ^/lo cm., is slightly smaller than the eye. Further, in L. fernande- 

 ziana there is no row of teeth behind the large conical teeth in the mandible, 

 while there occur an outer row of very small teeth labially of the great ones. 



The evidently closely allied L. rhacinus Forst. from New Zealand has 

 only six rays in the ventral fin, in which respect it agress with the related 

 L. viarginata Macleay (Proc. Linn. Soc. N. S. Wales, VI, 1881, p. 114. Not 

 of GUNTHER 1878, cfr. below.) from South Australia, and with L. bacchus 

 Forst. from the New Zealand coasts as well. The number of rays in the dor- 

 sal and the anal fins is perhaps also smaller in the new species, FORSTER and 

 HuTTON state for L. rhacinus the fin formula D. 5/68, A. 62, whereas the 

 same in L. fernandeziana (when Steindachner's statements are added) is 

 D. S/58-65, A. 55-60. 



In the Ann. and Mag. Nat. Hist., vol. II, 1878, p. 9 GiJNTHER described 

 a Lotella from the Pacific Coast of south-western South America, L. marginata 

 (see also Chall. Rep., Zool. XXII, p. 86). This species has the fin-formula 

 D. 7—8/65, A. 62, V. 5 and thus is more different Worn L. fernandeziana ihdin 

 the species mentioned above. 



As the specific name marginata already was given to a species of the 

 genus Lotella by GuNTHER 1878 (1. c), MacLEAY's name marginata of 1881 

 (1. c.) cannot be accepted. As apparently it has not been changed later, I 

 propose the new name Lotella macleayi for L. marginata Macleay 1881 (as 

 described in Proc. Linn. Soc. N. S. Wales, VI, 1881, p. 114 on specimen from 

 Port Jackson, Australia). 



3. Trachichthys fernandezianus Giinth., Rep. Chall. Exp., Zool., vol. 

 XXII, 1887, p. 23 (Juan Fernandez). 



One specimen, 89 mm. (s. c). Near the shore in a depth of 20 m.; 

 Masatierra ^"/s 191 7. 



The present sample agrees entirely with STEINDACHNER thorough de- 

 scription. 



