PISHES AND FISITINO IN SUNAPEE LAKE. 71 



these, at least, difTfrs from some other frosh-watcr sincUs as much as 

 it docs from the marine smelt, it seems more likely to confuse than to 

 clear up the matter to transfer and aj)i)ly names indiscriminately 

 without sufficient data upon which to hasc conclusions. 



The two ai)parent fresh-water races, previously alluded to, may be 

 only apparent on account of this same lack of knowledjj;e. The ap- 

 parent differences are those princii)ally of size and habits and to some 

 extent structure, so far as examination of specmiens has jiroceeded. 

 In a number of Maine lakes there are (apparently) two distinct sizes, 

 with somewhat different breeding and feeding habits. One size reaches 

 a length of at least 15 inches and a weight of a pound, and even larger 

 ones have been reported. The smaller one existing in the same lake 

 seems not to reach a larger size than 5 or 6 inches at most, as indicated 

 by those constituting the breeding runs in the streams. The time of 

 spawning differs more or less. The height of the period of the larger 

 form being at least a month earlier than the smaller one. The smaller 

 one usually ascends the streams as soon as they are free from ice, or a 

 little later. The larger one is known to ascend them, in some localities 

 at least, before the ice is out. 



In those lakes where there is apparently such an extreme difference 

 in size, only the larger form is caught mth hook and line in summer 

 and through the ice in winter, this being due to the difference in feed- 

 ing habits, the large smelt subsisting mainly upon smaller fish, for 

 the most part young smelts and the smaller form. The small smelt 

 subsists, so far as at present ascertained, almost wholly upon minute 

 crustaceans. This characteristic feeding habit obtains, however, only 

 where the two apparently widely distinct forms exist, as in some lakes, 

 Sunapee for instance, even little smelts only 4 or 4^ inches long are 

 taken on worm and fish bait. 



Then, again, there are lakes and ponds where the two distinct sizes 

 do not seem to exist and the smelts are of ])ractically a uniform size 

 in the one pond, differing in size variously from those of other ponds, 

 according to the pond; and some of the ponds are closel}'^ connected 

 with lakes in which the two sizes exist, others are far remote from 

 other ponds with smelts. 



Some of the large and deep lakes contain only tinj^ transi)arent 

 smelts, sexually mature when only 2 or 2^ inches long; while in a 

 neighboring body of water, at least within the same county, a much 

 smaller pond contains smelts 6 or 8 inches long. Also there is an 

 instance of a very large lake containing the two apparent extreme 

 sizes, with a tributary pond, the connecting stream of which is not 

 over one-half mile long but at present obstructed by a dam, in which 

 the smelts are uniformly of from 2 to 3 inches in length and sexually 

 mature. Thus it appears that the smelt question is at present a very 

 puzzling one, especially regarding their specific identity, and they 



