42 DROSERA ROTUNDIFOLIA. Cuar. IIL. 
the channel of communication between the two. On 
the other hand, such connecting threads are some- 
times seen to break, and their extremities then 
quickly become club-headed. The other sketches in 
fic. 8 show the forms successively assumed. 
Shortly after the purple fluid within the cells has 
become aggregated, the little masses float about in a 
colourless or almost colourless fluid; and the layer 
of white granular protoplasm which flows along the 
walls can now be seen much more distinctly. The 
stream flows at an irregular rate, up one wall and 
down the opposite one, generally at a slower rate 
across the narrow ends of the elongated cells, and so 
round and round. But the current sometimes ceases. 
The movement is often in waves, and their crests 
sometimes stretch almost across the whole width of 
the cell, and then sink down again. Small spheres of 
protoplasm, apparently quite free, are often driven by 
the current round the cells; and filaments attached 
to the central masses are swayed to and fro, as if 
struggling to escape. Altogether, one of these cells 
with the ever changing central masses, and with the 
layer of protoplasm flowing round the walls, presents 
a wonderful scene of vital activity. 
Many observations were made on the contents of the cells 
whilst undergoing the process of aggregation, but I shall detail 
only a few cases under different heads. A small portion of a 
leaf was cut off, placed under a high power, and the glands 
very gently pressed under a compressor. In 15 m. I distinctly 
saw extremely minute spheres of protoplasm aggregating them- 
selves in the purple fluid; these rapidly increased in size, both 
within the cells of the glands and of the upper ends of the 
pedicels. Particles of glass, cork, and cinders were also placed 
on the glands of many tentacles; in 1 hr. several of them were 
inflected, but after 1 hr. 35 m. there was no aggregation. Other 
tentacles with these particles were examined after 8 hrs., and 
