Lake, Kennebec County, southwest Maine (= Cobbosseecontee Lake which drains via Pleasant 

 Pond into the Kennebec River). The identity of this species is less certain. Its size, 

 4 inches (= 102 mm probably SL) , is closer to the size of the pygmy smelt than the rainbow 

 smelt. Other features such as scale number suggest the rainbow smelt. The type specimens 

 of Osmerus abbottii could not be located in the Academy of Natural Sciences by Dr. James A. 

 BShlke (in lit.). 



Kendall (1914) listed 3 species of smelt for Maine: Osmerus mordax , Osmerus spectrum and 

 Osmerus abbot! (sic). Jordan and Evermann (1896) and Jordan, Evermann and Clark (1930) 

 treated Cope's Wilton Pond smelt as a subspecies of Osmerus mordax . Hubbs (1925) in a 

 revision of the family, recognized Osmerus mordax north of New York on the Atlantic coast 

 and 0. sergeanti Norris, 1868 south of New York. He omitted from his study Osmerus spectrum 

 and CK abbottii which were being critically studied by W.C. Kendall (Kendall's taxonomic 

 study was never published). 



Kendall (1927) provided a valuable summary of the literature on the biology of eastern 

 North American smelts and added much unpublished information of his own and others. He 

 described the differing sizes of several populations but did not age specimens or report on 

 other morphological differences. 



Greene (1930) studied the smelts of Lake Champlain and commented that he could find no 

 morphological differences between the large and small races. McAllister (1963), in his 

 revision of the family, naively accepted Greene's verdict, and did not examine small smelt. 

 He divided Osmerus into Osmerus e. eperlanus of Europe and 0. e. mordax of the western 

 Atlantic, Pacific and Arctic, according subspecific status at least until sympatric 

 populations of the two forms were found not to interbreed. Klyukanov (1969) in his revision 

 of the genus reported sympatric non-interbreeding populations in Arctic U.S.S.R., applied 

 the name 0_. eperlanus to the European smelt, 0. m. mordax to all eastern North American 

 smelt, and distinguished Pacific and Arctic populations as 0. m. dentex; he did not discuss 

 0. spectrum . 



Zillox and Youngs (1958) studied the smelt in Lake Champlain and divided them into a 

 large race and a small race. A winter specimen of each shown in a photo were 233 mm and 

 158 mm TL. The angler's catch averaged 84% of the large and 16% of the small smelt. 

 Comparisons of growth of large smelt in 1929 and 1950 showed little substantial change; of 

 10 age samples compared, only two differed significantly. Thus growth within a "race" 

 in this lake appears relatively constant. 



Rupp and Redmond (1966) summarized the results of valuable experimental transplants of 

 six established smelt populations into eight new lake environments. They reported changes 

 in growth rate and longevity between source and transplanted stocks. They found that 

 "little" smelts became "big" smelts when transplanted to a reclaimed (rotenoned) lake and 

 concluded the effect of genotype was small in comparison with the potential influence of the 

 environment; thus implying that big and little smelt populations were largely environmental 

 in origin. The early spawning habit did appear to be under genetic control, however. 



Copeman and McAllister (1978) re-analyzed the data of Rupp and Redmond (1966) plus some 

 more recent data on these transplanted populations. Copeman and McAllister (1978) 



