1912] GRIGGS—RHODOCHYTRIUM 137 
sometimes as small as 12. Moreover, it is sometimes very 
tardy in its development. The variation in shape is likewise great 
(figs. 15, 17, 18, 20, 21, 24). LLAGERHEIM’s material shows for the 
most part the same sort of solid plugs. The form shown in fig. 19 
was observed but twice, while bell-shaped plugs such as he 
figures were entirely absent from that portion of his material 
which I examined. It seems safe to assume that the apparent 
discrepancy is to be explained by the same sort of variation as that 
just noted in the form on Ambrosia. 
There seems, therefore, to be no course open but to conclude 
that there is no morphological basis for separating the three forms. 
There is, on the other hand, reason to expect that they would be 
found to be physiologically differentiated in respect to their hosts 
if a series of experiments in cross infection were undertaken with 
one or all of the forms. Until the experimenter acquires more 
skill, however, than is possessed by the writer in transferring the 
infection from plant to plant,- the expected negative results of 
cross-infection would prove nothing. 
The range of the plant seems to call for some comment, but 
the data are hardly sufficient to decide whether the three known 
localities represent points in a single extensive range, or whether 
they are isolated stations. If they represent the continuous range 
of a single species, the limitation to such unrelated hosts raises 
some considerable difficulties concerning their distribution. Two 
of the hosts, Spilanthes lundi and Asclepias pumila, are somewhat 
localized species, and their range in neither case extends to 
either of the other stations; but Ambrosia artemisiifolia is wide- 
spread, and occurs both in Kansas and throughout South America. 
If the three forms are not physiologically distinct, therefore, 
cross-infection should occur naturally in Ecuador. 
It seems, therefore, that the answer to the question of the 
number of species of Rhodochytrium will depend on the point of 
view of the student. He to whom geographical and physiological 
isolation are criteria of species may well conclude that there are 
three species, while he who demands morphological characters by 
which to distinguish species will decide that there is but one. 
Each of these points of view has its advantages, and it is not for 
