158 THE AMERICAN MONTHLY [Sept 



the first of August and saw that it was changing into 

 Melosira horreri and that there was in many times to quan- 

 titatively appeared a form that I was not pre[»ared for. 

 The genus Cyclostella was present but no species that 1 

 could identify. Itwas nearly transparent but looked at very 

 closely it had fine rays or dots near the periphery of the 

 disc. They were all alive and had rays diverging from 

 them several times as long as the diameter of the f rustule. 

 These rays were of some substance which was not 

 cellulose for it was too hard for that, and seemed to con- 

 tain a certain amount of silica in its composition. For 

 when the gathering was dried on the cover aud ignited 

 over a spirit lamp, it did not entirely burn away. Tliis 

 was an old discovery, for I had seen the same thing 

 happen in a gathering I made ten years ago on the 

 marshes of Elizabeth, N. J. 



What was new was the presence of an amoeba. It was 

 nearest to Amoeba radiosa, at least it looked like Leidy'.s 

 figure 4, Plate IV, Fresh- water Khizopods. I say it does 

 not agree with Amoeba radiosa entirely for Leidy says it 

 is "comparatively small, colorless, transparent, inac- 

 tive." This was colored, light brown color and active. 

 But what is most strange is this. It seemed to come from 

 and be evolved from a Cyclotella. I saw a Cyclotella 

 with its rays which I have described. I saw another 

 Cyclotella with the rays more short, only just beyond the 

 siliceous frustule. I saw another in which the surround- 

 ing cellulose was growing more and more until it assumed 

 the form of an amoeba, and this was not an amoeba 

 eating a cyclotella, as I have seen hundreds of times in 

 years gone by. Soon the amoeba grew in dimensions, 

 and by and by put forth a portion which separated 

 from the parent and became a separate amoeba, moving 

 slowly away. After a time it slowly stopped and then 

 assumed tlie form which Leidy has figured in his figure 



