G. S. Graham-Smith 
47 
The first experiments were made with surface cultures on shallow 
layers of agar in Petri dishes. In the case of some organisms of the 
“slipping” group great differences were noticed in the types of the 
colonies formed on different parts of the dish. On searching for a 
reason of this phenomenon it was noticed that where the layer of agar 
was thicker owing to a depression on the bottom of the dish a certain 
type of colony developed, and where it was thinner another type. The 
different rates at which layers of various depths dried seemed to afford 
the most satisfactory explanation of the variations which had been 
noted. Owing to the uneven construction of these dishes the attempt 
to follow up these observations in them was abandoned and ultimately 
the following method was adopted. 
Thoroughly clean, sterile glass plates, made from old quarter plate 
negative glasses, were placed on the bench in a slightly tilted position 
so that one end was about 0'3 to 0'4 cms. higher than the other. 
Melted agar was poured on to the plate near the upper end and allowed 
to run down nearly to the other end. In order to secure an even layer 
it was distributed with a hot, sterile platinum needle. Several plates 
were made at a time in this way and allowed to solidify and dry under 
bell jars for various periods of time. Finally the emulsion of the 
organism to be investigated was applied usually in three or more 
parallel longitudinal streaks with a thin platinum loop. The plates, 
separated by supports, were then arranged horizontally one above the 
other in a large glass Petri dish. A capsule of water was put into the 
dish, the lid put on, and the whole incubated at 37° C. 
In the “ loop forming” group no important differences in the colonies 
produced on the thick and thin portions of the agar were noticed in 
thinly sown cultures. Usually the colonies were larger and more 
inclined to spread at the thicker end, but the general type remained 
the same. When thickly sown the colonies coalesced to form a continuous 
film in the thicker portions. Even in continuous growths however 
indications of the original colonies could be discerned. In the thinner 
parts the colonies were however separate. In all cases the edges of the 
growth, whether in the form of a continuous film or as separate colonies, 
showed the characteristic regularly looped condition, and extensions, 
such as are illustrated in Plate IV, Fig. 13, were common. 
The “ folding ” group behaved in the same way. The margins of 
the growths were however usually formed by thick wavy strands 
composed of parallel bundles of bacilli, and thin expansions of irregular 
loops, as shown in Plate V, Fig. 10, were uncommon. 
