1921] A Pure Culture Method for Diatoms 125 



Diatoms, as well as bacteria, have, in some cases, well defined 

 contour of colony. Each of the species we have cultured shows a 

 decided difference. The first type is that of a spreading form. It 

 soon comes to cover the entire plate with a film of individuals. A 

 plate of this form is shown in fig. 1. A colony of this type is easy to 

 secure since one has but to dip down between the colonies in an old 

 plate and make cultures from this "dip." Another form of colony 

 is shown in fig. 2. Here we find the colony margin to be restricted 

 and the form more or less radiate, with the organisms rather evenly 

 distributed over the area. A third form of colony is somewhat 

 similar to the latter, but differs in that the central area is much more 

 thickly settled than the margin. This thickened area occurs before 

 the gradual spread as is easily seen from fig. 3. Perhaps the more 

 characteristic form -is that assumed by the last type which, we call 

 the sheaf type. Fig. 4 is of this type. 



We are satisfied that we have not in any degree studied all the 

 forms that may be cultured in this manner, since a number of forms" 

 were found in the first plates which we did not have time to follow 

 up, and further, our original pond water did not contain a great 

 number of forms. 



Diatoms cultured in this manner are easily cleaned and prepared 

 for examination. The various colonies are spaded out, placed in a 

 test tube and the agar dissolved in boiling water. The solution is 

 centrifuged with a small centrifuge and the precipitate is washed 

 several times with hot water, the centrifuge being used each time 

 for concentration. After all the agar has been removed the Diatoms 

 may be either burned upon the cover glass or cleaned with sulfuric 

 acid and bichromate. After thorough washing they are kept in 50% 

 alcohol. 



The pure culture methods open up several fields of work. First, 

 the physiology of a species may be studied as was done by Richter. 

 Second, the classification of the groups may be studied. We think 

 this last point one of great interest. It is fairly well known that 

 species have been made upon the description of a single valve. By 

 this method, if the species will grow on agar, both shells would be 

 available for study and any differences could be noted. Rare forms 

 may be secured. Again, there is probably great variation among 

 the Diatoms, as elsewhere, and probably the majority of these variants 

 would show up in these cultures, thus species could be minimized. 

 If there should be any doubt as to the common ancestry of all the 



