GREAT LAKES COREGONIDS 



451 



usually is present on the cheeks, particularly on the oculars. The tip of the mandible 

 is pigmented but seldom is conspicuously dark. The preorbital region is like the 

 cranium in respect to pigmentation, and the maxillary always shows more or less of 

 pigment. All the fins are pale, but the caudal and dorsal are darkest. These show 

 a more or less faint dusky hue on their distal margins. The pectorals often 

 show faint pigment on their longest rays; the anals sometimes have a few dots 

 on the membranes between the rays, but the ventrals, except in very rare cases, are 

 immaculate. 



During the breeding season pearl organs are developed by males and by at least 

 some females. Their development apparently is not very different from that described 

 for johannse. However, there are occasionally one or two much smaller pearls 

 flanking the central one of the scales of the scale rows of the sides, and on the scales 

 of the rows dorsad and ventrad to the fourth above and the sixth below there are 

 regularly two or three or even more pearls on each scale, the disparity in size decreas- 

 ing as the back and belly are approached and the distribution and shape becoming 

 more irregular. 



VARIATIONS 



Racial variations. — It will be seen from Table 56 that a considerable num- 

 ber of specimens has been collected from almost every port visited. The lowest 

 number from any locality is 8 from Platte Bay, and from all but 5 of the 16 

 other stations from which specimens were preserved 34 or more specimens have 

 been examined. These collections are fairly uniform in respect to size of individuals, 

 with the exception of the Michigan City and Northport lots, which have a greater 

 proportion of large specimens. Compared in all their important systematic charac- 

 ters, as number of gill rakers on the first branchial arch and of scales in the lateral 

 line, and the relative size of head, eye, and paired fins, there are no differences dis- 

 cernible between the various groups except such slight ones as might be due to in- 

 equality in size of the various individuals composing the groups, namely, changes 

 in proportion that are the result of growth. 



There is another possibility of racial differentiation, namely, according to 

 habitat. In Lake Huron, for example, it has been observed that specimens from the 

 deepest water differ in certain characters from their shore relatives (seep. 458), but 

 in Lake Michigan it has not been possible, from the collection I have accumulated, 

 to establish any such differences. My specimens, however, do not lend themselves 

 to any such comparisons, as they were collected over a period of several of the warm- 

 est months, and it is known that the bloaters move nearer shore at certain seasons, 

 so that a given habitat in the same locality might be occupied by different races 

 during a season. A study of environmental races, then, must be undertaken first 

 in a definite and restricted area over a period of time. For the present all that can 

 be said about variation is that the collection of about 1,000 individuals from 17 

 stations scattered along the lake's shore does not disclose any striking variation 

 tendencies. 



Size variations. — By far the greater number of specimens collected are less than 

 200 millimeters in length, and the largest ones are but little over that limit. In 

 Table 57 five specimens over 200 and five under 200 millimeters in length are com- 

 pared extensively, and in Tables 8 to 11 all the specimens over 200 millimeters in 



