248 University of California Publications. [Botany 



sharply defined upon which to work, and it was selected as a rep- 

 resentative form for a complete and careful study, after which it 

 was used as a basis for comparison with similar structures in 

 other plants. First a number of standard aqueous nuclear stains 

 were used upon the living material. The plants were washed in 

 alcohol and examined in clove oil (just as when stained with 

 methylene blue), with the result that the same structures uni- 

 formly appeared. Among the stains used were ruthenium red, 

 neutral red, rubin G., safranin, toluidin blue, china blue, and 

 thionin; and of these the best in general were found to be the 

 blue stains, since they act like double stains, coloring the granules 

 red, thus differentiating them from the other structures, all of 

 which stain blue or bluish. The red stains do not so clearly sep- 

 arate the granules from the thread-like structure in the center of 

 the cell. 



To supplement and check this work, plants were killed with 

 several different killing agents, e.g., alcohol, Flemming's solu- 

 tions, 1 per cent, chromic acid, and potassium iodide-iodine, after 

 which the stains mentioned were applied. Without exception the 

 same structures were revealed as before, but not equally well de- 

 fined. Chromic acid and Flemming's solution followed by the 

 blue stains do not yield so clear and definite results as these stains 

 do on the living material, nor are the blue stains so good after 

 these killing agents as the red stains. Material killed with potas- 

 sium iodide-iodine or with alcohol stains about the same as the 

 living material. 



Corroborative evidence coming from such a variety of sources 

 as we have here warrants the conclusion that the structures re- 

 vealed by the methylene blue on the living material are real and 

 not artifact. Furthermore, the selective action of these chroma- 

 tin stains as indicated here is evidence that the thread-like struc- 

 ture is composed of chromatin. 



If in all of the species of Cyanophyceae the chromatin were 

 as easily differentiated as it is in Symploca Muscorum, little trou- 

 ble would be involved in the interpretation of its behavior during 

 cell division, but this is found not to be the case. The abundant 

 granules in the center of the cells of some species, as well as 

 granules embedded in the peripheral cytoplasm, have been a 



