18 ANDERSSON, COMPARISON OF COTTUS POECILOPUS WITH COTTUS GOBIO. 
My measurements in this respect show the following. In 
five out of 22 small specimens (30—50 mm. in length) of 
C. gobio this measurement was nearly the same as the cor- 
responding average in C. poecilopus and out of 22 specimens 
of middle-size (50—70 mm. in length) five showed the same 
slight difference from the average of C. poecilopus as the first 
ones, beside which one specimen of C. poecilopus out of 15 
showed a still closer resemblance to C. gobio. In the largest 
specimens (70—95 mm. in length), in which the differences 
ought to be more marked, there were 5 or 6 out of 24 spe- 
cimens of C. gobio, which, to judge by this measurement, 
might be almost classed amongst OC. poecilopus. Moreover, 
out of 20 specimens of C. poecilopus four displayed a marked 
likeness to OC. gobio. 
It appears from this that it is impossible to use this cha- 
racter as a trustworthy and specific one, although, on the 
whole, there exists a certain difference in this respect between 
the two forms. 
Nor can the other character which Professor SMITT gives 
— the different length of the base of the anal fin — be re- 
garded as a significant distinction between the species. Ac- 
cording to Professor SMItTT's diagnoses, the length of the fin 
is in C. gobio less than 48 2 of its distance from the tip of 
the snout and in C. poecilopus more than 50 2 of the same 
distance. It may well depend on our different ways of mea- 
suring, mentioned above, that of 34 males measured of C. go- 
bio I found no less than 18 in which the length of the fin 
exceeded this measurement, and of an equal number of fe- 
males 6 such. Professor SMITT's too low measurement may 
also be due to the circumstance that he measured one male 
only and four females, of which latter the measurement is 
always less than that of the males. Of 51 specimens of C. 
poecilopus measured by me, one only did not come up to 
Professor SMITT's minimum-measurement, but the belly of this 
one was distended by large ova, which caused the beginning 
of the anal fin to be pushed backwards, for which reason the 
distance between the anal fin and the tip of'the snout was 
greater. It seems at any rate impossible to fix any other 
measurement marking in this respect the distinction between 
the forms. According to my examination in 13 specimens out 
of 68 of C. gobio the relative measurement was greater than 
