FISHES AXn I'ISIIINC IX SUNAPEK LAKE. 77 



(Infinite ovcler, somotinios downstream, soinetiinos uj), ami sonictiiucs 

 cross \\'iso, and often some headinj]; in one direction and some in anotlier. 

 In another- ])Ool above tliis a school occupied an eddy, swimminfij al)out 

 irrejxnlarly and sknvlv to some extent, and <2:eiierally rath(T stationary 

 or driftiiiij: iiTe{j:ularlv, but Antli their heads }2:enerally toward the slow 

 return current at jdmost ri^j^ht any;les to the bank. 



In another pool a school started by the writer's step on the i>ank 

 darted downstream as far as a shoal ripple, then slowly retm-ned with 

 heads all directed upstream, some smelts above others, but all in the 

 same direction. The smelts when undisturbed did not all occu]iy tho 

 same level in the water; some were near bottom and some farther uj) 

 in the water, even at times near the surface, but they were all the 

 time rising and settling again, s\vimming back and forth individudly 

 and to some extent collectively but irregularly in the latter case. 

 Th(>re was no e\'idence that they were at this time spawning. In the 

 first pool mentioned a few eggs wei;e seen attached to dead leaves, 

 moss, and sticks, but they were white and may have been extruded 

 when the fish were disturbed the ])revious night by dip{)ing. Further 

 observations show that the smelts very slowl}' movc^d about in the 

 eddy in a comparatively large "circle" or rather ellipse, but in a very 

 iiTegular manner. 



Two smelts, one large and the other smaU, were seen to come rather 

 quickl}^ to the surface together, breaking water with their backs. 

 Probably this was not significant, as no more were seen to do it, or 

 anything like it, during a long watch. No evidence of pairing was 

 observed. 



Later in another place a small school of smelts was seen lying at the 

 foot of a j)ool in which was considerable current. They were compara- 

 tively motionless, just above a shallow ri])ple, heads all upstream, 

 merely (h-ifting from side to side, when with one or two quick flirts of 

 the tail they kept themselves from going backward. They scarcely 

 moved upstream at all at any time, and when there was such a move- 

 ment it was only on the part of one or two of them, not the whole 

 school. 



At 9 p. m. the smelts had mostly gone out of the deej) holes and 

 were scattered along the brook, generally on the ripples, but on the 

 morning of April 17 the schools were all in the deep holes where 

 they were seen during the day before. 



On the night of April 18 the writer observed some smelts in the 

 brook by the hatchery that were evidently spawning, making no 

 attempt to go farther up the brook. There were, however, others 

 above and some running uj) by them. Those watched were in shallow 

 water on sand, fine gi'avel, and pebbles and headed uj)stream where 

 the current ran (puckest, but nearer shore they would lie on the 

 bottom with their heads in no particular dhection. Sometimes 



