juneis, igis Azotobacter Chroococcunt 237 



present writer; perhaps because he did not use the same methods as 

 Penau. As can be seen from the accompanying illustration (PI. XXXI, 

 fig. 6, 7) the granules do not disappear from the cells at any time of the 

 life cycle of the latter; neither do they show any uniform setting. Pos- 

 sibly they would correspond in A. chroococcum, to the so-called nuclei 

 of B. anthracis if they did not exist already in the undifferentiated 

 stadium and if they did not persist in the reticulated stadium. As 

 Penau states, the metachromatic granules already exist in the undiffer- 

 entiated cells of B. anthracis and persist in the reticulated cells, while 

 the structure which he calls true nucleus disappears at the outset. 



In Plate XXXI, figure 6, No. 6, and in Plate XXXII, it can be seen 

 that the granules may even be placed on the outside of the cell. IMencl 

 (11) also has noted this fact, but does not attempt to give any explana- 

 tion in regard to it. The granules increase in size with age, but, aside 

 from this character, they present no other. 



Their plurality could not by itself exclude their nuclear nature, but 

 that character in addition to their occasional extracellular position 

 would be sufficient to deny to them nuclear functions. This point was 

 not considered by Mencl as one worthy of attention. 



How could he consider these granules, escaping from the cells as 

 nuclei, as the most important and most vital organs of a cell? It is true 

 that in higher plants some cells (sieve tubes) are also left in old age 

 without nuclei, but it is also true that in most cases these cells are no 

 longer capable of reproduction. These granules have also been found to 

 escape from cells which were undergoing the process of division. No 

 isolated mass of cytoplasm has ever been seen to divide spontaneously. 



Plate XXXI, figures 8 and 8a, shows some other cells stained by the 

 Guignard (6) method. All stages are here represented. 



Some cells, which are easily found, do not contain any granules, 

 although their size indicates an advanced age. This lack of granules 

 might possibly be attributed to an expulsion of the same by the method 

 just mentioned. 



To judge from Mend's (11) drawings, many cells present one granule, 

 but no regularity as to its setting in them. In Plate XXXII, which repre- 

 sents some cells drawn at random from a Guignard-stained preparation, 

 the granules are always seen embedded in the cytoplasmic matrix ^ and 

 are never to be seen inside the meshes of the cell network. Sometimes 



1 The expressions "cytoplasmic matrix, " "cell network, " "cytoplasmic strands" are here used to mean 

 that part of the cell contents that in our organisms has an affinity for basic dyes. 



As we have already seen, the Romanowsky compound stain gives a differentiation of colors in the cells, 

 the network taking the blue color characteristic of nuclei and the contents of the meshes taking a pink color. 

 Some comparative preparations with normal and abnormal human blood were studied; the same staining 

 solution as the one used for A . chroococcum gave the ordinary colors. 



Nothing should now prevent the naming of the basophylic cell constituents "nucleo substances," and 

 the eosinophylic ones "cytoplasm. " But since we mean to furnish more proofs to establish their nature, 

 we will continue to use the terms "matrix, " "network, " "strands, " according to their morphology. See 

 also Plate XXXI, figures 2, 3, and others. 



