20 



the genial autumn and spring would have been most beneficial,, 

 and the extreme seasons the most baneful. But so it is not. 

 The hot, dry, dusty summer is the most propitious of all, and 

 next our winter, with its sub-tropical rains. 



Now it may be suggested that since this is a chronic and not 

 an acute disease, although during the autumn the mortality is 

 highest, this may be the effect of the excessive heat of summer ; 

 the effect not being perceptible until after the lapse of some 

 months, because the heat does not kill the phthisical patient at 

 once, but only after an interval. This suggestion is plausible, 

 and moreover it derives no litttle support from the figures, for 

 we might thus explain that it is not only the great heat of 

 summer that produces the great mortality of autumn, but that 

 the milder autumn weather accounts for the smaller deathrate 

 of winter, the comparative severity of winter again the 

 increased mortality of spring, and the genial influences of 

 spring the minimum mortality of summer ; allowing thus three 

 months for the production of the effect by its cause. But 

 although its ingenuity is attractive, this theory must not, I 

 think, be entertained. Bather than receive it, we must lay the 

 statistics aside, with the impression that we can come to no 

 certain and definite conclusion about the influence of season. 

 For how do we know what length of time must elapse before 

 the deleterious influence acting on the phthinode will kill ? 

 Why should it be three months, and not six ? Why should we 

 suppose that the higher mortality of the autumn is the 

 expression of the baneful influence of the summer, manifested 

 after an interval of three months, rather than that the lower 

 mortality of the winter is the expression of its beneficial 

 influence visible after six months ? No ; unless we conclude 

 that the small summer deathrate is due to the summer, we 

 must abandon all conclusion on the effect of season. Did we 

 find in reference to phthisis a state of things similar to what 

 we observe in considering the question of infant mortality, we 

 might admit the hypothesis perhaps. There we see the death- 

 rate increasing vert/ rapidly during the hot months of December, 

 January, and February, to a high figure, but reaching the 

 maximum only in March, when the temperature has begun to 

 fall, and then decreasing as the heat continues to fall during 

 the succeeding months. In such a case it is admissible as an 

 explanation, and highly probable, that the maximum in March 

 is the climax of the ill effects of the high temperature of 

 December, January, and February, and is not due to the 

 March temperature alone. But the case is very different in 

 regard to consumption. If the summer influences were dele- 

 terious, we should at least expect that during the summer the 

 deathrate would increase to some extent, even though it should 



