10 ANDERSSON, COMPARISON OF COTTUS POECILOPUS WITH COTTUS GOBIO. 
Very often we meet with specimens of C. gobio with spotted 
ventral fins, but in these the spots are not usually so re- 
gular nor so distinct as in C. poecilopus. We find, however, 
in C. gobio in these spots a tendency to take the form of 
transverse bands, this being the case in a few of the speci- 
mens I examined; and, as the spots in OC. poecilopus also are 
sometimes indistinct, it would not, this being the case, be 
correct to pronounce them distinct species, judging by this 
character alone, which, even without regard to intermediate 
forms, seems to be of comparatively little importance. 
C. gobio seems, in this respect, to have departed least 
from the primitive type, probably corresponding pretty ex- 
actly with the parent form, from which C. poecilopus differs 
in having a larger amount of the pigment in the fins. The 
cause of this difference is very difficult to ascertain; but it 
may perhaps depend partly on local circumstances. The two 
forms have probably begun to differ at different places, though 
they afterwards spread into each-other's localities. 
In the specific diagnoses given by HEcKEL in his later 
work he still retains the spottiness of the ventral fins as the 
chief character, but he no longer insists upon the different 
branching of the rays in the fins, introducing instead of this 
the following distinctions between the two species: The vent- 
ral fins are long in C. poecilopus and reach the vent, while 
they are shorter in OC. gobio and do not reach the vent; more- 
over the tail in OC. poecilopus is not so deep, only !/14 of the 
length of the body, while in OC. gobio it is from Y/12 to Y/13 of 
the length of the body. 
'The different length of the ventral fins. 
This character is also mentioned by Professor LiILLJEBORG, 
who gives it in his specific diagnoses, declaring it to hold 
good at least in specimens of any size. GÖNTHER too states 
it in his »Catalogue of Fishes», and JEITTELES says: »von speci- 
fischem Werth ist aber die fast gar nicht veränderliche Länge 
der Bauchflossen». It appears from the following table, show- 
ing the average length of the ventrals in proportion to that 
of the body. that there exists in the two forms a very re- 
markable correspondence in the length of the ventrals in 
