270 University of California Publications. [Botany 



stain one kind and not the other; and equally important, one 

 that would not stain the other cell structures to obscure the gran- 

 ules. Two stains were found that will do this unerringly in all 

 the species worked upon. These are Bismarck brown and was- 

 serblau. 



One kind of granules, which for convenience will be called a, 

 are located within the nucleus in every healthy living cell of each 

 species studied, with perhaps a few exceptions among the unicel- 

 lular forms and in the genus Merismopedia (which were investi- 

 gated at the beginning before proper technique was employed, 

 and which I have not since been able to obtain). There is an- 

 other exception which I have not yet been able to explain satis- 

 factorily, viz., mature spores. 



The location of these granules is variable and depends upon 

 the shape and size of the cell and also upon the structure of the 

 nucleus. They are always in close proximity to the chromatin, 

 but not imbedded in it, even in small filamentous species where 

 the chromatin is often in a single irregularly shaped mass (Fig. 

 38). In the large species of Oscillatoria and other large filamen- 

 tous forms they are scattered promiscuously through the nucleus 

 among the chromatin elements. 



The number of these granules is also variable in the same 

 species at different times and in different species. The variation 

 is probably due to differences in cell activity. In resting cells 

 there is a tendency toward a constant number. At least this 

 seems to be the case in small-celled species in which one can easily 

 count them. There are often filaments of Lyngbya Lagerheimii 

 (Fig. 38) in which there is but a single large granule close to 

 the chromatin mass in each cell throughout almost the entire fila- 

 ment. These a granules occur sometimes with such regularity as 

 to lead one to suppose them to be nucleoli, but they do not be- 

 have like nucleoli of higher plants, which tend to disappear dur- 

 ing cell division. On the contrary, they seem to increase in num- 

 ber when the cells are actively dividing. 



The exception above spoken of, i.e., that they do not occur in 

 the spores, seems significant. This was a difficult point to deter- 

 mine since the spores are not easy to stain on account of their 

 thick walls, yet some stains were found which penetrate the wall. 



