318 



BULLETIN OF THE BUREAU OF FISHERIES 



It appears that only clupeaformis and hoyi have a distinctive number of scale 

 rows — the one with the highest, the other with the lowest count. The range of both, 

 however, overlaps, more or less, that of at least some of the other species. Within 

 each species there may be variations in the number of scale rows in the various lakes 

 and among the several races within the same lake. Thus, johannsp has fewer scale 

 rows in Huron than in Michigan; zenithicus, reighardi, nigripinnis, and Jioyi have 

 fewer in Nipigon; hoyi has slightly fewer in Superior than in the other lakes; the pre- 

 dominant artedi of Erie, Ontario, and Nipigon have fewer rows than the predomi- 

 . nating herring races of the other lakes; and the Erie whitefish has fewer than the white- 

 fish of the other lakes. The manitoulinus subspecies in Huron has many fewer rows 

 than the artedi subspecies, and the alius subspecies of Erie has fewer than the artedi 

 form in that lake. This is true also of these two races in Lake Superior. It seems, 

 in general, that the number of scale rows varies directly with the number of lateral- 

 line scales. 



Pearl Organs 



Pearl organs are present during the breeding season in at least the males of each 

 species. Not enough material is available to determine to what extent this character 

 is of systematic importance to separate the species, but it appears that it will separate 

 the three groups. In Leucichthys pearls are present on the head as well as on the sides 

 of the body and are of virtually uniform thickness. They have been found well devel- 

 oped only in males. In Coregonus they are distributed more or less as in Leucichthys 

 but are present on males and females and are conspicuously thicker in the middle. 

 All the specimens of Prosopium seen differ from Leucichthys and Coregonus in having 

 no pearls on the head. The form of the pearl is approximately as in Coregonus, and 

 both males and females have pearls. 



Fins 



DOKSAL 



The ratio between the height of the dorsal and its base (the dorsal coefficient 

 of the tables) is variable in all species and even in all races. There are no distinguish- 

 ing features about this character, but it is interesting to point out that the dorsal base 

 appears to average longest, relatively, in the various races of clupeajormis, quad- 

 rilaterale, and artedi, and shortest, relatively, in hoyi and Myi. 



The number of dorsal rays also is not distinctive except possibly between the 

 forms with extreme numbers. Quadrilaterale, which has the highest number (11 to 

 13), overlaps but seldom the range of hoyi, with usually 9 or 10. Kiyi and reighardi 

 also usually have a low dorsal ray count. 



The number of rays seems to vary among the races of a species as well. Thus, 

 the Superior and Nipigon forms of reighardi have more dorsal rays, on the average, 

 than the typical form. Hoyi, in Nipigon, tends to have a greater number than the 

 forms elsewhere in the Great Lakes. 



ANAL 



The anal coefficient is variable and distinguishes only Prosopium absolutely 

 from most of the species of Leucichthys. The anal base is relatively shortest in 

 quadrilaterale, hoyi, and Jciyi and longest in johannx and artedi. In the Leucichthys 



