1 6 Davis. — Spore Formation in Derbesia . 
that of ascestral forms and can have no value in phylogenetic considerations. 
The evidence indicates that the ancestors of Derbesia produced immense 
numbers of small zoospores represented by the very numerous nuclei which 
enter the sporangium, and it is very probable that these zoospores were 
similar to the biciliate elements characteristic of the Siphonales. 
The zoospore of Oedogonium has a structure similar to that of Derbesia 
with respect to the arrangement of its cilia, and I venture to predict that 
the development of its blepharoplast will be found to parallel closely that of 
Derbesia , for the reason that the zoospore is developed from an exceptionally 
large uninucleate protoplast. However, I hardly think that any one for 
these reasons will be bold enough to suggest a relationship between 
Derbesia and Oedogonium in view of the striking differences at all points in 
their general morphology. 
Derbesia may serve as a warning of the danger of carrying too far the 
practice of basing classifications of the algae on the structure of the zoospores 
and motile gametes. The removal by some authors of Vaucheria from the 
Siphonales seems to me open to serious criticism, and has not been followed 
by other authors, as, for example, Oltmanns (’04). Botrydium is of course 
a peculiar type of uncertain position. This study of Derbesia shows clearly 
the necessity of treating exceptional forms of algae separately, deferring 
judgement on the significance of peculiarities of zoospores and motile 
gametes until their development and structure is understood in comparison 
with these elements in groups of supposed relationship. 
Summary of the Investigation of Derbesia. 
The protoplasm in the globular sporangium becomes separated from 
that of the parent filament by the formation of a heavy cellulose plate, 
which begins to develop as a ring at the periphery, and consists of layers 
deposited successively as the protoplasm in the sporangium and filament 
withdraw from one another. 
The protoplasm within the developing sporangium contains several 
thousand nuclei. These are all similar to one another, and slightly larger 
than the plastids at the time when the cross-partition is formed at the base 
of the sporangium. Each nucleus contains a small nucleolus and a large 
chromatin body. 
A process of nuclear differentiation sets in very shortly after the 
protoplasm of the sporangium becomes separated from that of the parent 
filament. Certain nuclei increase in size, reaching a diameter 4-6 times that 
of the plastids. These larger nuclei also become conspicuous because of 
the development of numerous protoplasmic strands which radiate out from 
the cytoplasm enveloping the nucleus. The strands have deeply-staining 
granules where they join the enveloping cytoplasm just outside of the 
