APPENDIX— CYCLOPIUM GYGLOPUM. 139 



a little behind a line with the posterior end of the anal fin. The third (1-9 

 long) has no appearance of an adipose dorsal spine. The smallest (TS long) 

 has a distinct spine with foui' spinate elevations along its upper surface, and 

 which are stronger than the similar elevations in the other fins. This series 

 is very interesting, for it demonstrates that a spine in the adipose fin may be 

 present or al)sent, and likewise that the spine if .present may l)e smooth or 

 spinate along its upper edge. While these differences alone may he ol)served 

 in these silumid fishes, which otherwise entirely resemble one another, they 

 may reasonably be considered to be identical species. 



Ten si^ecimens from the neighl)ourhoo(l of Machachi. One was examined 

 and found to contain 1.3-16 vertel>ra', while the air-bladder was enclosed in 

 bone as in the loaches. It would be tedious to follow out each specimen as 

 all agreed witli the before-descril)ed form. The five smallest from 0-9 inches 

 to 1'5 inches in length show the adipose dorsal spine Avith its spinate upper 

 edge. From 1-8 inches in length to 2-4 the long and low adi])ose fin is still 

 visible, but in another at 2'4 inches it has entirely disappeared. Six other 

 specimens taken from near the same locality and all of small size show the 

 adipose dorsal spine — this series evidently proving that not only may the 

 adij)ose dorsal spine hi atrophied in large examples but also that all trace of 

 the dorsal fin itself may be lost. 



Twenty -seven specimens were received from the neighbourhood of 

 Cayambe. The lengths of these fish \'arv fi'om 0-7 of an inch to 2'7 inches, 

 while most ai'e infested with parasites. Piecisely the same chai'actei's are 

 shown, liut the adijiose dorsal fin is not so completely lost in the larger ones 

 as in those from Machachi. This would seem to be due to the examples 

 Ijeing rather better prepared. Were these specimens kept in weak sjiirit 

 doiU)tless they would take on the appearance of a long, low adipose dorsal fin. 



As to breeding the ova are comparatively large, while tlie male appears 

 to possess an intromittent organ. 



My examination of the foregoing series of siluroids fit^m the Andes 

 induces me to coincide with Dr. Putnam's views. 1. That the adipose dorsal 

 spine has not even a specific value. Although all four specimens from Cliillo 

 agree with Arges brachycephalus, Giinther, so likewise do the larger examples 

 with SUjgogcnes Humboldtii, Gimther. 2. That the names of the fish have 

 the following priority. Cydo'pium (Swainson) cyclojmm, Hundjoldt : Pime- 

 lodus cydojnun, Hund)., 1803 : Cycloirium Humboldtii, Swainson, 1839 : Aryes 

 cyclopum, Guv. and Val., 1840 : Arges brachycephalus, Giinther, 1859 : Hfygo- 

 genes Humboldtii, Giinther, 1864 : and Stygogeni's cyclopum, Giinther, 1864. 



