STUDIES ON BOVINE MASTITIS                              367

for the following reasons. It should be noted that in 1932 the laboratory control
was made more strict and, also, that a selective enrichment medium was introduced
capable of detecting very small numbers of streptococci in milk. The considerable
increase in the number of new infections revealed in Herds A and C in 1932 may
thus be accounted for. It follows at the same time that the incidence of infection
prior to 1932 was probably higher than is represented by the figures given. Sub-
sequent to 1932 the number of new infections fell, although not so much as was
anticipated. In both herds a large number of new infections—mostly with non-
hæmolytic Str. agalactiæ—were diagnosed solely by means of the enrichment
medium and such animals were practically always giving normal milk. In a few
cases chronic infections developed later or an obvious mastitis associated with
large numbers of streptococci, but in nearly all cases infection was shown at one
test and not at subsequent tests. In practice, however, such cows were always
treated as infected and they appear as such in the tables, so long as they were
still present in the herd. Thus, although the infected sections appear large in
these two herds, many of the cows therein are not in fact infected at the present
date. On the other hand, most of the cases which were diagnosed by direct
plating of milk deposits have shown persisting infections. It should be said,
finally, that the streptococci concerned in the transient infections mentioned were
typical Str. agalactiæ biochemically and serologically and in some instances have
been found capable of producing a chronic infection when injected into the udder
of goats experimentally via the teat canal.

The experience in Herds A and C is of particular interest in establishing
the fact that Str. agalactiæ is capable of setting up transient self-healing infections
which can only be detected by the most stringent bacteriological tests. Since
most of such infections were diagnosed by the enrichment medium only, there
is some reason for regarding the use of an enrichment medium as superfluous
for the control of the disease in practice.

Herd L remains for comment. This, it will be remembered, was formed
from freshly-calved heifers and hence was expected to remain entirely free from
Str. agalactiæ. As already stated, however, one heifer was found to be infected
at her first test and two others subsequently became infected. Nevertheless,
the disease is well under control and, although it is still too soon to conclude that
infection will not appear later, the incidence of infection amongst these animals,
most of which are now in their second or third lactation, need only be compared
with that shown in Table 6 to realise that where chronic mastitis is at all prevalent
the formation of a separate heifer herd is a measure of the greatest value.
Emphasis has been laid in another article [Stableforth, 1935] on the practicability
of forming a separate herd from young animals and the value of this procedure
in protecting them from infection with other chronic contagious diseases besides
mastitis.