nally deficient rainfall. Again, the highest death-rate
registered since 1865 was in 1893, a year of similarly exceptional
drought. But it is no more than an hypothesis, and the fate of former
theories is a warning against drawing conclusions from statistics and
records extending over too short a period of time. The warning is
particularly necessary in connexion with meteorological conditions,
which are apt to upset all calculations. As it happens, a period of
deficient rainfall even greater than that of 1854-1858 has recently been
experienced. It began in 1893 and culminated in the extraordinary season
of 1899. The dry years were 1893, 1895, 1896, 1898 and 1899, and the
deficiency of rainfall was not made good by any considerable excess in
1894 and 1897. It surpassed all records at Greenwich; streams and wells
ran dry all over the country, and the flow of the Thames and Lea was
reduced to the lowest point ever recorded. There should be, according to
the theory, at least a very large increase in the prevalence of
diphtheria. To a certain extent it has held good. There was a marked
rise in 1893-1896 over the preceding period, though not so large as
might have been expected, but it was followed by a decided fall in
1897-1898. The experience of 1898 contradicts, that of 1899 supports,
the theory. Further light is therefore required; but perhaps the failure
of the recent drought to produce results at all comparable with the
epidemic of the 'fifties may be due to variations in the resistance of
the disease, which differs widely in different years. It may also be due
in part to improved sanitation, to the notification of infectious
diseases, the use of isolation hospitals, which have greatly developed
in quite recent years, and, lastly, to the beneficial effects of
antitoxin. If these be the real explanations, then scientific and
administrative work has not been thrown away after all in combating this
very painful and fatal enemy of the young.

Dissemination.

The conditions governing the general prevalence of diphtheria, and its
epidemic rise and fall, which have just been discussed, do not touch the
question of actual dissemination. The contagion is spread by means which
are in constant operation, whether the general amount of disease is
great or small. Water, so important in some epidemic diseases, is
believed not to be one of them, though a negative proof based on absence
of evidence cannot be accepted as conclusive.