g6 
F. E. Fritsch. 
ceased after a time and that then the adjoining vegetative cell 
became changed to take its place and so on. The fewer the 
heterocysts at the end of an old spore-producing filament the larger 
they generally are. When abundant spore-formation is going on, 
the cells next to the heterocysts are apparently often in an 
exhausted condition and if they divide, the results of their division 
are below the average size of the vegetative cells. Heterocyst- 
formation may also affect such cells and the resulting heterocysts 
are then also below the average size. As a rule when several 
heterocysts are formed in a row the first-formed ones tend to become 
disjointed. Long filaments, the centre portion of which is engaged 
in active spore-formation, and which are bounded by a number of 
heterocysts, not rarely exhibit a number of collapsed cells with 
faint blue-green homogeneous contents (necrides) next to the chain 
of heterocysts. In many cases the cells near the fully differen¬ 
tiated heterocysts are more or less homogeneous and contain few 
granules (cf. fig. 9.) 
The filaments of the Anabaena were readily stained with an 
aequeous solution of Iodine. The contents of the cells took on a 
dark brown colour almost at once, whilst even half-an-hour’s 
immersion in the stain had no effect on the heterocysts. Young 
heterocysts, which, as mentioned above, still have coloured con¬ 
tents like those of the remaining] cells, are also totally unaffected 
by the Iodine. Is this due to a change in the contents or to a 
change in the permeability of the cell-wall or to both ? However, 
not only the heterocysts remain uncoloured by Iodine; the above- 
mentioned collapsed cells, occurring frequently in the neighbourhood 
of the heterocysts, are also unaffected. I can find no grounds for 
assuming an alteration in the cell-wall to have taken place here, 
whilst the contents have visibly undergone a change. The action 
of an aequeous solution of methyl-green is almost directly the 
reverse. The contents of the heterocysts 1 in all stages of 
development take on a bright green colour, whilst the vegetative 
cells are far less prominently coloured; apparently chiefly the 
granules inside the latter take up the stain. The collapsed cells 
take up the stain in the same way as the contents of the heterocysts, 
thus again evincing a similarity to these latter. A short immersion 
1 When methyl-green is brought in contact with the smaller 
heterocysts it causes a dissolution of the membrane and 
only the terminal thickenings and the contents remain 
evident; in the larger fully-differentiated heterocysts the 
membrane tends to contract irregularly (cf. Brand, loc. cit. 
p. 39). The action of dilute acids is similar. 
