502 PROCEEDINGS OF THE ACADEMY OF 



The characters of the Trachinidse and Blenniidse given by Gunther are 

 essentially interchangeable, with the exception of the following : 



Trachinid.se. "One or two dorsal fins, the spinous portion being always much 

 less developed and shorter than the soft ; the anal similarly developed as the soft 

 dorsal ; ventrals with one spine and five rays.* Gill openings more or less wide."\ 



Blenniidse. ' ' One, two, or three dorsal fins, occupying nearly the whole of 

 the back, the spinous portion, if distinct, being as much developed as the 

 soft, or more." " Ventrals jugular, composed of a few rays, and sometimes 

 rudimentary or entirely absent." 



Only two " cardinal characters" have thus been used to distinguish the 

 Trachinidse and Blenniidse. 



Dactyloscopus was said by Gunther to have " one dorsal, formed by spines 

 only ;" it therefore had nominally the distinctive characters of the Blennioids 

 as understood by that gentleman. 



I have, on the other hand, specifically asserted that only the first eleven 

 or twelve rays are spines, the others (22 31) being "articulated, and divided 

 on each side of the mesial line to the base, but so connected as to appear like 

 simply articulated rays, especially from a lateral view." Gunther's observa- 

 tion is therefore incorrect. % 



Dactyloscopus then agrees with the Trachinoids and diners from the Blen- 

 nioids in a character which has been emphatically insisted upon by Dr. Giin- 

 ther, and to which the structure of the ventrals has been always subordinated 

 by him. 

 " It disagrees with the Trachinoids and agrees with the Blennioids in the 

 structure of the ventral fins ; a character which Giinther has elsewhere 

 regarded as of little importance. 



It therefore, according to Gunther's diagnosis, only differs from the Trachi- 

 noids in one "cardinal character," which is of much less value than the 

 cardinal character which it shares in common with the Blennioids. 



Further, it agrees with the Trachinoids and departs from the Blennioids by 

 the width ot the gill openings, and also differs from the Blennioids by the 

 large scales. 



Accepting Dr. Gunther's own views of the relative value of certain charac- 

 ters, Dactyloscopus is thus more allied to the Trachinoids than to the Blen- 

 nioids. Therefore, it was probably only on account of a misapprehension 

 that the genus was referred to the Blennioids. I shall, however, still refer to 

 the arguments adduceable in favor of its reference near the Uranoscopoids. 



The form of the head of a Blennioid is quite characteristic, owing to the 

 abrupt decurvature of the profile in front of the eyes, and the almost or quite 

 horizontal cleft of the mouth. 



Equally characteristic is the form in the Uranoscopoids, the profile in front 

 of the eyes being continued on nearly the same plane as the crown, while the 

 cleft of the mouth is very oblique or vertical. 



Dactyloscopus agrees in general form with the Uranoscopoids. 



The Uranoscopina3. Leptoscopinse and Dactyloscopinse agree with each and 

 differ from the Blennioids in 



1st. General form. 



^ 



* In Epicopus with one spine and Bix rays. Giinther. 

 f The italicized parts are repeated from Gunther's Work. 



I I am happy to state, that Dr. Gunther has since admitted that the anterior rays of Dactylos- 

 copus alone are spinous. In a letter of the 25th May, he writes : " Your statement of a portion 

 of the dorsal rays being articulated is correct; they are very well preserved in the smallest of 

 our specimens, (18 lines long,) whilst in the larger (30 lines) most of them are broken at the top, 

 a I now see." Dr. Gunther has not given his present opinion of the affinity of the Dactyloscopi. 



3 It is proper here to remark, that the Uranoscopus adhaesipinnis of Blyth (Journal of the 

 Asiatic Society of Bengal, vol. xxix. (I860,) p. 42,) does not belong to the same family as Uranosco- 

 pus but apparently belongs to the same genus as the Polycauhis elongatus (Giinther ex Cuv.) 



[Oct. 



