184 Dr. H. Charlton Bastian on Diatoms and the 
of the thallus (a comparatively large Navieula and a Cocconezs) 
are never found within the substomatal spaces or the epithelial 
cells. 
3. Chemotaxis implies a direct power of movement in 
response to an attractive influence; but none of the diatoms 
on the surface of the duckweed within the spaces or within 
the epithelial cells have ever been seen to move. 
4, The diatoms in the spaces are found intimately inter- 
mixed with the algoid cells, and generally in situations to 
which they could not be supposed to have the power of 
penetrating. 
5. The diatoms can often be seen to have replaced algoid 
cells rather than to have pushed them aside. 
6. Finally, in places, the algoid cells can be seen elon- 
gating into the forms of the diatoms and at the same time 
changing from a bright green to a yellowish-brown colour. 
Moreover, since making these observations on L. gibba and 
L. minor I have ascertained that similar transformations of 
some of the fission-products of the Chlorochytrium which 
infests LZ. triésulca are also to be met with in that species of 
duckweed. ‘The diatoms found in this species have been 
almost always very small and of the Navicula type—no 
Niteschie having ever been seen in association with the 
segmentation-products of this particular variety of Chloro- 
chytrium, although the duckweed bearing it has been taken 
from one of the same ponds from which I have obtained my 
supplies of L. gibba and L. minor. In fig. 6 (x 3875) some 
of the combinations that have been met with are shown. In 
A four small spaces are shown: in the upper one algoid 
segments and diatoms were intermixed; in the one on the left 
young diatoms were seen forming, the contents of this space 
being distinctly paler than those of the other two spaces, in 
which the diatoms were more fully formed and more closely 
packed. In B two or three fused contiguous spaces are shown 
in which algoid cells and diatoms, together with various inter- 
mediate forms, were intimately intermixed; while C is the 
only space that I have yet found in ZL. trisulca containing 
diatoms as large as are there represented. ‘They were mixed 
with Chlorochytrium cells and other minute diatoms, though 
the latter are not recognizable in the photograph. 
It is worthy of note that in Lemna trisulca there are no 
stomata. The active algoid spores penetrate, as F. Cohn 
showed, by boring between the epithelial cells into subjacent 
spaces, where they increase and multiply in practically closed 
cavities and become also surrounded by a kind of capsule. 
Subsequently their active spores make their way out through 
ee 
