GREAT LAKES COREGONIDS 



331 



size of the customary food organisms is contingent and secondary and the gill- 

 raker characters have the greater value in taxonomy. The differences in form and 

 position of the bones of the upper and lower jaws, as well as other characters ex- 

 hibited by the different forms, may ultimately prove to belong in the same category 



Experimental study of the effect of environmental factors on fishes in their 

 various stages, together with breeding experiments, are, of course, essential to a full 

 understanding of the characters shown by the coregonid forms. The production, 

 by artificial rearing, of very abnormal characteristics in the whitefish indicates that 

 this field may be very fruitful. Meantime, more critical analysis of the accumu- 

 lated data on variation may throw light on these problems. 



The only conclusion that can be drawn safely from a consideration of the varia- 

 tions in these forms is that only certain characters are modifiable by environmental 

 conditions or tend to vary, and that these characters are virtually the same for all 

 the forms. Thus, the variants of all species may differ from the typical forms in 

 respect to head length, number of scales in the lateral line, length of pectorals, num- 

 ber of gill rakers, etc. Furthermore, so many characters are variable that the vari- 

 eties may be more strikingly differentiated from their nearest relatives than are the 

 species within a group from one another. 



SPECIATION IN THE GREAT LAKES COREGONIDS 



ORIGIN OF THE COREGONIDS IN THE BASIN 



In this paper I have presented evidence to show that 11 distinct species and 

 7 possible subspecies of the family Coregonidse are found in the Great Lakes chain. 



There are 9 species included in the genus Leucichthys and 1 each in the genera 

 Coregonus and Prosopium. Lake Michigan has 10 of the 11 species, Lake Huron 9, 

 Lakes Superior and Nipigon 8 each, Lake Ontario 7 (though 1 now seems to be extinct), 

 and Lake Erie 2. The distribution of the various species in each lake is shown in 

 Table 5. In general there is little difficulty in correlating the relationships of the 

 various species in the different lakes, but in a few cases the individuals of a species 

 have varied so far, structurally and even physiologically, from the typical form as to 

 appear to merit designation as a distinct species. There is no doubt about the rela- 

 tionships within the species artedi (the most variable of all) nor within the species 

 johannse, alpense, zenithicus, Jciyi, hoyi, clupeaformis, or quadrilaterale. In the 

 case of reighardi it might be questioned whether the dymondi form of Lakes Nipigon 

 and Superior actually should be included within that species group, and in the case of 

 nigripinnis whether the cyanopterus of Lake Superior and the prognathus of Lake 

 Ontario were grouped properly within nigripinnis. The case of the dymondi and 

 cyanopterus races seems the more confused, as, in addition to marked structural differ- 

 ences (though the habitat selection is about the same), the time of spawning is very 

 different. In this connection it may be pointed out that time of spawning may vary 

 two weeks from year to year for any school of any coregonid, and the time of spawn- 

 ing of races of many species (namely, clupeaformis within Lake Michigan, zenithicus 

 of Lake Huron, Tciyi of Lake Ontario, et al., all of them of virtually certain identifica- 

 tion) may be a month or two earlier or later than for related races elsewhere in the 

 basin. It is even reported that within recorded time certain species have changed 



