148 The Tj/pe o/Exocoetus exiliens [L. OmeL). 



count tlie rays, but tlie former seems to be formed of 12 or 

 ]3, and the latter of 9 or 10 rays. The dorsal Jin commences 

 Jar in advance of the anal, in fact, tlie first anal ray is opposite 

 to the sixth of the dorsal ; on the other hand, the basis of the 

 anal extends rather more backwards than that of the dorsal. 

 Tlie dorsal is elevated throughout its length, some of the 

 hinder rays reaching nearly to the caudal. The lower caudal 

 lobe in its present dried condition is considerably longer than 

 the head (33 nim.). The majority of the scales are lost, but 

 there seem to have been 42 in a longitudinal series on the 

 side of the body, and 28 between head and dorsal fin ; I 

 count 9 longitudinal series above the first anal ray. 



Tlie hinder |iart of the dorsal, possibly the entire fin, was 

 black; the anal whitish, tlie extremity of the caudal blackish. 

 Basis of the pectoral blackisii, then follows a broad white 

 cross-land from the lower to tiie upper margin ; the posterior 

 half of the fin black, with traces of a narrow white margin. 

 Anterior half of ventrals whitish, posterior blackish. No 

 black band across the abdomen. 



The characters given above tally very well with the notes 

 left by Gmelin, except the numbers of the fin-rays, which 

 Gmelin states to be D. 10, A. 11. However, no great weight 

 can be attached to this discre|)uiicy, considering the frequent 

 inaccuracies which we meet in this respect in Gmeliii's edition, 

 as well as the great difficulty in ascertaining tiie correct 

 numbers in this particular specimen, the fins of which must 

 have been always (since it came into Linne's possession) in 

 the same collapsed and dried-up condition. Mr. Tate Kegan 

 assisted me in fixing the numbers by counting the swelled 

 joints at the base of the fin-rays. 



Gmelin says : " pectorales radio primo et secundo brevi- 

 bus " ; we must therefore conclude that he saw and counted 

 the rudimentary ray which I find on one side of the specimen, 

 but not on the other. But this condition is very different 

 from that obtaining in the young individuals which Liitken 

 (Vid. Meddel. 1876, pp. 110 et seq.) determined as £, exi- 

 liens, and which have that anterior ray much more developed, 

 nearly one half the length of the second. 



Jordan and Evermann (Fish. N. Amer. p. 732) ascribe the 

 priority of the original description and the name '* exsiliens'^ 

 to P. L. S. Lliiller, giving 1776 as the date. This is an 

 error. That date is the year of publication of the^/.s/ volume 

 of Miiller's work : the description of our fish appears in the 

 seventh volume (Supplement) dated 1789, a year later than 

 Gmelin's edition of the ' Systema naturge,^ and published 

 several years after Miiller^s death. Besides, Miiller (or, after 



