212 Cholera in Relation to Certain Physical Phenomena. [part i. 



Next, comparing the first and third columns, we find an entire agreement for the 

 first six months ; but after this, December and November change places with February 

 and May, whilst March and April again occur in reversed order. The principal use 

 of the second comparison is to show how, in examining the results of very short 

 periods, whilst we find a general agreement with those of longer ones, minor differences 

 are introduced by the exceptional occurrences of particular years. 



Eeturning now to the consideration of the tables as a whole, it will be found that 

 the agreement between their results may be considerably augmented by removing 

 the figures regarding one great cholera season — that of 1865-66 — from the second 

 table. If we do this, starting with September 1865, in which the great rise in 

 prevalence began, and removing the figures of each of the subsequent months up to 

 the following September, we get the following series of figures as the monthly totals 

 for the remaining 1 1 years : — 



TABLE XI. 



Total GholeTct Deaths from 1865 to 1876, exclusive of the 'period between 

 September 1865 and September 1866. 



The tables now agree, save that June and July change places with October and 

 August. April now appears in both tables as the month of maximum prevalence, and 

 the relation which the figures for November and December bear to one another comes 

 to be closely similar in both cases. 



Let us now see what the average annual course of cholera as exhibited in these 

 figures really is. Taking November as most nearly representing the average prevalence 

 of the disease, and therefore as a good starting-point, we find successive diminutions 

 in prevalence during December and January, a rapid rise in February, continuing to 

 the maximum in March and April, a marked diminution in May, continued through 

 June, to a minimum in July, August and September, and finally a rise in October to 

 reach the average in November, Stating the facts of prevalence in most general 

 terms, it may be said that there are four months in which the prevalence of cholera 

 greatly exceeds the average, three months in which it falls far short of it, and five 

 months in which it ranges round it, the prevalence in November approaching it more 

 nearly than that of any other month. 



Such are the results of the more important masses of data at our disposal, and 

 we may next examine those relative to various limited communities in order to see 

 how far they agre^^ with the others. 



