668 PROCEEDINGS OF SECTION H. 
There being practical unanimity among authorities about the 
importance of local insanitary conditions, such as defective drain- 
age, it is clearly of the utmost importance that there should be 
definite knowledge about the way in which these conditions 
operate. Direct contagion being relegated to a quite subordinate 
place, we have left as possibilities the following, as modes of 
conveyance of the virus :— 
I. Contaminated water supply. 
II. Accidental contamination of articles of food, and 
especially milk. 
III. Inhalation of emanations from the soil, or from cess-. 
pits, drains, &e. 
That typhoid not infrequently spreads through the medium of 
contaminated water may be taken as proved. The most commonly 
recognised mode of contamination is by the washing, into wells, 
channels, or reservoirs, of the virus contained in the fecal dis- 
charges of persons suffering from the disease. It is also probable 
enough that not fresh discharges, but virus which has been 
developed in the soil or in drains, soaks into wells or other recep- 
tacles ; and by many it is held that it is only in this indirect way 
that an undrained, filth-saturated soil favours great prevalence of 
typhoid. This may readily enough be the case in small towns 
and rural districts, where the water supply is derived from 
streams or wells liable to impure soakage, and where the intro- 
duction of a single case may supply a source of infection to soil 
or water, or both, and so lead to quite an extensive outbreak. 
But it is evident that the conditions are altogether different in a 
town or city deriving its water supply from some outside source, 
from which it is conveyed in pipes. Even then, of course, there 
are two sources of danger—there may be contamination at the 
head-waters ; or there may be occasional suction into the pipes 
locally, where the supply is intermittent. In the case of Mel- 
bourne it can be safely affirmed, I think, that the sources of its 
water supply are better guarded than those of most large cities. 
As the population in the gathering area is small, specific contami- 
nation could take place only in an accidental way, to a slight 
extent, and at comparatively long intervals. 
Assuming the possibility, or even the probability, of this 
specific contamination at the sources, it is apparent, however, that 
its result would be a sudden outbreak of limited duration. The 
annual outbreaks of the disease, with their beginnings in Novem- 
ber, their steady increase till February or March, and their regular 
decline during the late autumn and winter, are not explicable at 
all on this assumption. The regularity of the rise and fall is 
shown on the accompanying chart. The possibility of local con- 
tamination by the suction of foul matter into the pipes, from the 
soil or from the street channels, must also be admitted. But 
