Fission-products of a Parasitic Alga. 183 
almost all that has been said against the infection hypothesis 
and in favour of the transformation hypothesis as accounting 
for the presence of the diatoms in the substomatal spaces 
holds good also in regard to their presence in the epidermal 
cells. In one respect the argument is even stronger in its 
application to them, since there is much evidence to show 
that diatoms are only found in those epidermal cells which 
are or have been tenanted by the alga, and such infected cells 
never constitute more than the smallest fraction per cent. of 
those existing on the whole upper surface of a leaf. 
A further point of extreme importance is to be found i in the 
very great differences in the size and shape of the diatoms, 
according as they originate from the small or the larger 
algoid fission- products. Yet these variations, for which no 
ether contributory cause is apparent to us, are 80 great that 
botanists unaware of the origin of the diatoms, and finding 
them in the Chlorochytrium spaces, would almost certainly 
regard some of them as belonging to different species of the 
same genus, and others even as representatives of distinct 
genera. This, however, is a subject which must be left for 
future investigation. 
It was suggested to me by a distinguished botanist, to 
whom I showed some of the specimens of duckweed containine 
in their substomatal spaces and epithelial cells mixtures in 
various proportions of Chlorochytrium segments and diatoms, 
that their association might be explained by the Infection 
Hypothesis, backed by the assumption that Ckemotaxis had 
been in operation, which in this case would mean that the 
physico-chemical processes associated with the growth and 
multiplication of the alge within the spaces were capable of 
giving rise to products exercising au attractive influence upon 
the diatoms. 
It was not pretended that there was any direct evidence in 
favour of this assumption; it was advanced as a possible 
explanation and merely to stave off the conclusion, otherwise 
inevitable, that the diatoms had been produced by. the trans- 
formation of the cells of the alga. 
A careful and unbiassed consideration of the following 
facts will, however, I think make it plain that the evidence 
is overwhelmingly against Chemotaxis and the Infection 
Hypothesis :— 
1. Chemotaxis can only be supposed to operate at short 
distances; but such diatoms as are found within the spaces 
are never to be seen on the surface of the duckweed. 
2. The diatoms that are commonly met with on the surface 
