4 BULLETIN OF THE 



Here they were also kept on the diet of mosquito larvoe until from one 

 cause or another they had all died. 



My failure to secure early stages of the eggs in the spring of 1882 

 made me desirous of repeating the attempt at a more seasonable time 

 the following year. With this object in view I left Cambridge for 

 Ogdensburg, May 18, 1883. 



Judging from my previous experience that it would be difficult to 

 procure fertilized eggs in sufficient quantities without great labor, if they 

 were to be individually detached from the rocks, I procured several yards 

 of thin muslin of a color resembling the stones in the lake. I planned 

 sinking this and loading it with small stones in the water on some of 

 the "points" most frequented by the gar-pikes at spawning time. I 

 hoped in that way to secure a large number of eggs firmly attached to 

 the cloth, which I could then remove to a box suitably provided with 

 wire nettings to allow the necessary circulation of the water. Had it 

 proved successful, this device would have enabled me to have under 

 control the eggs thus acquired, and would have allowed me a degree of 

 certainty as to the age of the preserved material not otherwise easily 

 attainable. 



Unfortunately for my plans, the weather proved to be in several ways 

 very unpropitious. A long period of cold and rain delayed the spawn- 

 ing to a time much later than common, and when at length, a few days 

 after the 1st of June, the weather and water became warm enough to 

 impel to the act of spawning, such high winds prevailed that it was 

 impossible to watch the movements of the fishes, and the most of them 

 had spawned before the water became quiet enough to allow one to dis- 

 cover their places of rendezvous. Some of the localities which they had 

 visited with the greatest constancy during the past years were appar- 

 ently deserted. Moreover, the cloths, which had of necessity been an- 

 chored near the shore, were either set free by the dashing of the waves, 

 or rolled into ropes which presented a comparatively small surface for 

 the reception of ova. 



The limited time in the latter part of May left at my command, after 

 completing some work which I was compelled to take with me, was util- 

 ized in studying the ovarian ova of females captured before the spawn- 

 ing period arrived, and in some attempts at artificial fertilization. I 

 then succeeded in getting a fairly satisfactory knowledge of the interest- 

 ing structure of the e^ membranes and of the micropyle, but I did not 

 learn the peculiar relation of the latter to the granulosa until a few 



