VoL - 2 J Gardner. — Cytological Studies in Cyanophyceae. 249 



source of much confusion, and it was not until I was able to dif- 

 ferentiate the granules and stain each kind separately without 

 staining- the chromatin, and further to stain the chromatin with- 

 out staining the granules, that this source of error was removed. 



The stains I found to be particularly troublesome in failing 

 to differentiate granules from chromatin and in yielding uniform 

 results on different species are Heidenhain's iron haematoxylin 

 and Flemming's triple stain. Early in the work Ehrlich's hae- 

 matoxylin was tried without good results, but the poor results 

 were found later to be due to the condition of the stain, which 

 was probably either "unripe" or "over ripe." Later some new 

 Ehrlich's haematoxylin was made up, in which Grubler's haema- 

 tin was used instead of haematoxylin. The mixture was placed 

 in the sunlight and occasionally shaken for a few days, until the 

 haematin and alum were mostly dissolved. 



This stain was found to be the very best differential chroma- 

 tin stain of the entire series used. It has the advantage of being 

 an excellent killing agent, so that the living material mounted on 

 the slide may be killed and stained at the same time, and has the 

 further advantage of differentiating well after a variety of kill- 

 ing agents. With this it is possible to stain the chromatin with- 

 out staining any of the granules. To do this requires but a few 

 minutes. If left in the stain longer, the central granules also 

 absorb it and become even more deeply stained than the chroma- 

 tin. Over-staining may be washed out with acidulated alcohol, 1 

 per cent, to 2 per cent. HC1 in 95 per cent, alcohol. 



II. RECENT LITERATURE. 



That the structure of the Cyanophyceae cell has been exceed- 

 ingly difficult to work out and to understand is evidenced by the 

 fact that from the time the task was begun until the present there 

 has been continuous disagreement concerning the nature of the 

 various parts of the protoplast. 



It is not the purpose of the writer to go extensively into the 

 literature on the subject in this brief paper. Several good re- 

 views have been published from time to time, among which are 



