146 
little creature assumes something of a pear shape, and from the 
narrow end a whip-like thread, the flagellum, is produced, at the 
base of which lies the nucleus. The flagellum is indeed, as Dr 
Jahn has shown, a continuation of the nuclear substance. The 
swarm-cell now swims about with a jerking or dancing move- 
ment caused by the lashing of the flagellum, and searches for 
food. I well remember the astonishment with which | first 
saw a swarm-cell feed. I had by me a specimen of Stemonitis, 
whose spores hatched remarkably quickly, so that within an hour 
after placing them in water under a cover-slip scores of swarm- 
cells were dancing under one field of the microscope. It 
happened that I had on the table, a wine glass in which a woody 
fungus had been soaking for a week or more, and the water was 
turbid with hosts of bacteria. Out of curiosity I put a drop of 
the turbid water to the edge of the cover-slip. Immediately the 
bacteria spread with rapid movements among the swarm-cells, 
and then I observed that most of the swarm-cells had thrown 
out a brush of delicate pseudopodia from the thicker end, oppo- 
site to the flagellum, and to many of these pseudopodia 
struggling bacteria were attached; these were gradually drawn 
into the body substance, where a vacuole formed round them. | 
have often watched the process since, and seen several bacteria 
enclosed in a digestive vacuole, and during the course of an hour 
or so have observed them to be gradually dissolved and entirely 
assunilated. Swarm-cells may often be seen having two or three 
digestive vacuoles crammed with bacteria. I have watched two 
swarm-cells, that had seized a bacterium longer than their own 
breadth by the opposite ends, contending for 20 minutes which 
should gain possession, with fluctuating advantage until one 
gave way and the bacterium was swallowed up by the victor. 
Besides the dancing motion, the swarm-cells creep in a linear 
form with the flagellum extended in front. I once watched one 
of these that came in its travels upon a group of bacteria on a 
glass plate; the vibrating tip of the flagellum appeared to detect 
their presence, for the swarm-cell stopped and spread itself as a 
disc over the cluster; in about a quarter of an hour it again 
crept forward with three of the bacteria stowed away in a diges- 
tive vacuole, and soon the swarm-cell gave a_ lash with its 
flagellum, released itself from the surface of the glass, and went 
dancing away among its swimming companions. 
A day or two after the swarm-cells have hatched they begin 
to increase in number by bipartition ; the flagellum 1s eae 
the swarm-cell takes a globular form, and the nucleus divides by 
karyokinesis. In the course of about a quarter of an hour the 
two halves of the nuclear plate separate and the globular ae 
cell becomes ellipsoid; then a constriction takes place, the ce 
divides, and each half resumes the form of the parent, and puts 
