94 ° 
Journal of Agricultural Research 
Vol. V, No. 20 
During the period of extensive multiplication, the plates from sandy 
loam, clay loam, and muck showed quite similar types, although the 
sandy loam has slightly greater numbers of the strongly aerobic spore- 
fonning species. As the numbers diminished, spore-bearing types 
became more frequent on plates from both sandy loam and clay loam, 
but were not evident on the plates from muck. It is to be inferred that 
the multiplication of those in the finest soil had not progressed to such 
an extent as to make their colonies numerous in high dilutions, their 
numbers apparently being in proportion to the grain size and amount of 
aeration. The most persistent non-spore-bearing organism was of the 
type already referred to, as found in clay and sand. In addition to this, 
certain chromogenic cocci and one variety of slime-forming organism were 
frequent on plates from all three of these soils through the time of desicca¬ 
tion. This slime former, which was especially numerous on plates from 
muck, is described as follows: 
The organism is a rod 0.4^ by 0.6 to 0.7^; nonmotile. No spores observed. No 
capsule demonstrated. Stains readily with aqueous alcoholic fuchsin. Nutrient 
broth made slimy and very turbid. Growth on agar spreading, translucent, orange- 
yellow, slimy. Gelatin stab, surface growth and rapid liquefaction. Litmus milk 
discolored, alkaline, slimy; peptonization begun in 48 hours and complete in 10 
days. Facultative anaerobe. No indol from Dunham’s peptone solution. Ammonia 
produced from Dunham's solution and nitrates reduced. Habitat, soil. 
Attention should be called to the rather peculiar circumstance that not 
one of the organisms isolated during the last two months corresponds to 
any one of the four organisms which predominated in the original soil 
solution used for the inoculation of the five soils. The extinction of these 
species jnay have been due either to the unfavorable influence of associa¬ 
tion with other organisms during the period of active multiplication or to 
their lack of endurance when supplied with less than the optimum 
amount of moisture. 
CONCLUSIONS 
(1) The survival of non-spore-bearing bacteria in air-dry soil is due, in 
part, to the retention by the soil of moisture in the hygroscopic form. 
This, however, is not the only factor, for the longevity of bacteria in a soil 
is not directly proportional to its grain size and hygroscopic moisture. 
(2) Bacteria, at least those tested, resist desiccation longer in a rich 
clay loam than in sand, under the conditions of our experiment. 
(3) If bacteria are suspended in the solution extracted from a rich day 
loam before being subjected to desiccation in sand, they live longer than 
if subjected to desiccation after suspension in physiological salt solution. 
(4) The solution extracted from a rich clay loam contains substances 
which have a protective influence upon bacteria subjected to desiccation. 
