Gr. S. G-raham-Smith 
49 
Where the layer of agar is thick and moist a film tends to be formed. 
In a thickly sown plate each bacillus produces a colony in the manner 
illustrated in Plate VII, and by the fusion of these colonies a fairly even 
film results. If the plate is less thickly sown isolated colonies are_ 
formed from each bacillus. From these processes run out, and join 
with those from neighbouring colonies forming a network. The 
uncovered spaces are filled in by lateral growth from the fused 
processes. The exact condition ultimately reached depends however 
on the rate at which the surface of the agar dries, or is in some other 
way rendered unsuitable for the development of further growth over 
it. If this occurs before the spaces of the network are filled in a 
discontinuous film as shown in Plate VIII, Figs. 8 (lower part) and 9 
(upper part) results. 
Isolated colonies growing on partially dry agar sometimes assume 
beautifully regular star-shaped forms. Plate V, Fig. 11 (nat. size) 
shows an extremely fine example of a colony entirely consisting of 
branched processes, and Fig. 12 illustrates the same colony after a 
further period of 24 hours’ incubation. Here it can be seen that very 
little lateral extension has taken place, but the processes have enlarged 
especially at their extreme ends, where large rounded masses of growth 
have been produced. 
In the driest part of the agar the growth is most limited as lateral 
expansion is early arrested, and rounded or oval heaped up colonies 
are generally formed. For a certain time the condition of the agar 
allows of lateral growth, and during this period a flat colony is formed. 
Subsequent growth results in the bacilli becoming superimposed and a 
thick colony being produced. Growth under these conditions has been 
studied and is illustrated in Plate VIII, Figs. 1—4. In Fig. 1 an 
elongated colony consisting of parallel bacilli all lying in the same 
plane is seen. Fig. 2 shows further development. In Fig. 3 it is 
seen that continued multiplication has caused some of the bacilli in 
the centre of the colony to override others. In Fig. 4 the difference 
between the central and peripheral portions is still better marked. 
The central portion consists of at least two layers of bacilli and the 
parallel arrangement is lost. The peripheral portion still consists of a 
single layer of bacilli. At this stage a curious phenomenon was 
noticed, namely very rapid motility at intervals on the part of groups 
of bacilli lying in the peripheral parts of the colony. These move¬ 
ments often began and ended rather suddenly and the bacilli never 
passed beyond the original limits of the colony. The blurred appearance 
Parasitology m 
4 
