28 Olive, Mitotic division of tlie nuclei of tlie Cyanopliyceae. 
granules seem to be in tlie cytoplasm, and to lie ontside tlie 
bluisli, irregularly defined central body. In fig. 94, which shows 
a fnlly matured spore of tlie sanie species, similarly stained with 
metliylene blue, tlie dense onter cytoplasmic zone seems, on tlie 
contrary, to be free from granules, which now appear wholly 
witbin tlie central portion. Fig. 93 , stained with iron haeraatox- 
ylin and eosin, is a longitudinal section of a very young 
spore, in whicli tlie cyanophycin granules are stained red, and 
are seen to be located only in tlie cytoplasm. Fig. 100 is a 
cross section of a half matured spore, showing large and abun¬ 
dant red granules, also in the cytoplasm alone. Figs. 101—103 
are cross sections of young spores, all stained with Flemming's 
triple stain, and all showing a well defined resting nucleus: 
while figs. 96, 97, and 99 are similar sections of old spores, 
similarly stained. Finally, fig. 95, a median longitudinal section 
of a mature spore, should be rioted. A careful comparison of 
these figures leads us irresistibly to the conclusion that while. 
in the young spores, the nucleus appears to begin to enter upon 
a normal resting state, in the older spores, the abundant cyano¬ 
phycin granules have encroached so upon the middle, sap-filled. 
nuclear cavity, tliat they are finally forced into it and fill the 
nuclear space. Figs. 96, 97, and 99 all show clearly the un- 
stained, globular spaces in which lie the cyanophycin granules, 
some of which in the two latter figures are located still in the 
cytoplasm and otliers witliin the nucleus. In figs. 95 and 96, 
all the food granules appear to lie witliin the limits of the 
nucleus. In the preparations from which these latter drawings 
were made, the chromatin is stained dark reddisli or purplisli, 
so that there can be no mistäke as to the identity of the rainute 
nuclear granules which are seen in the interstices between the 
unstained cyanophycin bodies. 
A possible explanation of this peculiar plienomenon seen in 
the spore of Cylindrospermum is afforded by the density of the 
periplieral protoplasm, although this is a point of which I have 
not yet convinced myself. As the cyanophycin granules accu- 
mulate in the cytoplasm in the i mm erbäte neighborhood of the 
nucleus, they finally become so abundant tliat they are probably 
forced, on account of their increasing numbers as well as on 
account of the density of the protoplasm in which they lie, 
into tlie nuclear cavity. They thus break down the delicate, 
forming membrane, push in among the chromatin granules, and 
in this way, in the mature spore, present the curious appearance 
of an enormous central body, which is completely filled with 
reserve food granules. We can see, moreover, in figs. 95—99. 
that the cyanpliycin occupies only the outer portion of the 
central body, while tlie middle is filled with a poorly defined, 
achromatic substance, in which are imbedded cliromatin granules. 
Such a peculiar encroachment upon nuclear space has certain 
resemblances to the phenomena seen in the spores of Nostoc 
(figs. 59, 60), in which the nuclei appear to be pressed into 
