book background

"Cholera, and the water supply in the south districts of London"

View this document as a PDF

British Medical Journal
(17 October 1857): 864-65

PDF courtesy of PubMed Central, NIH's free online digital database.

By John Snow, M.D.

A passage, respecting the effect of water supply on cholera, in the recent document published by the General Board of Health, having been brought prominently before the British Medical Association, in a leading article of the Journal of October 10th, I think it highly desirable to point out that this passage does not by any means convey the whole truth, as regards the effect of the water supply on the epidemic of 1854. Instead of the cholera mortality in the houses supplied by the bad water being 3½ times as great as in the houses supplied by the better water (the statement of the Board of Health), it was in reality six times as great.

It may perhaps be asked, of what consequence are the exact proportions, so long as the principle is admitted? If the report were merely intended to produce in public authorities and private individuals a certain amount of scrupulosity with regard to the supply of drinking water, the exact numbers would perhaps not matter much; but, when the facts are laid before medical men, it is of the utmost importance that they should be correctly stated. Science cannot be advanced by incorrect quantities and numbers; and, in the present instance, the real facts have an important bearing on the question as to the nature of the material in the impure water which induces cholera, and the manner in which the same morbid material causes the disease in other cases, without the aid of water as a medium.

My attention had been closely applied to the particulars of the water supply of London for upwards of five years before the epidemic of cholera of 1854; and I had, from various sources, became [sic] acquainted with a number of circumstances which made it possible for a very conclusive personal inquiry to be made, in respect to that kind of influence of water supply on cholera which I had published in 1849. These circumstances were necessarily known to a number of workmen and several official persons, but probably not in their collective form to any other person interested in the mode of propagation of cholera, except myself. The particular circumstances of the water supply, and its adaptation to the kind of personal inquiry which I conceived and undertook, were first published by me in the Medical Times and Gazette of September 2nd, 1854, p. 247, and were alluded to in a leading article in the Association Medical Journal for October 27th of the same year.

I called myself at every house from which a cholera death had been registered, in the first seven weeks of the epidemic of 1854, in all the districts in which the supply of the two Water Companies in question was intermixed; and, if the illness had not commenced in the house in which the death took place, I then sought the real place of attack, and in either case I ascertained the water supply of the house. I did not rest satisfied with a mere verbal reply; but obtained, in all cases, such corroborative evidence as could leave no doubt on the point, and I have the notes of my result. As a proof of their general correctness, I may mention that Mr. Greenwood, the very intelligent registrar of Christchurch, Southwark, made an inquiry on the same point for the same seven weeks, in his district, and, on our comparing notes, our results were exactly the same in every instance, although our respective inquiries had been conducted in a different manner.

The result of my inquiry was that, in the first four weeks of the epidemic, the cholera was between thirteen and fourteen times as fatal in the population having the impure water supply of the Southwark and Vauxhall Water Company, as in the population having the improved supply of the Lambeth Company, taking into account the number of the population supplied respectively by each company. In the next three weeks of the epidemic, the mortality was nearly eight times as great in the one population as in the other.

Dr. Farr having been much struck with the results which I communicated to him of my inquiries, the Registrar-General, at the end of these seven weeks, directed the district registrars in the south districts of London, to furnish the water supply of each house in which a fatal attack of cholera might occur during the rest of the epidemic.

During this part of the epidemic, comprising ten weeks, and including its most severe period, the mortality was still more than five times as great in the population supplied by the first of the above mentioned companies as in that supplied by the other; so that the result of that part of the inquiry conducted through the office of the Registrar-General, afforded a strong corroboration of the correctness of the previous part. I have, moreover, shown in an article in the Journal of Public Health for October 1856, that the whole of the inquiry agrees with the relative mortality of the different districts and sub-districts supplied in varying proportions by the two Water Companies, both at different periods of the epidemic, and for the whole epidemic, in such a manner as could not happen unless the results of the inquiry were substantially correct. I have already stated that the relative mortality of the two populations differently supplied with water was six to one, when the whole epidemic is considered.

The results of the above inquiry having been canvassed by the Scientific Committee of the General Board of Health, a further inquiry was instituted, and was carried out within the eighteen months following the epidemic, and furnished the numbers quoted in the Journal of the Association. This further or supplemental inquiry was conducted as follows. Lists of the houses supplied by each Water Company were obtained from the two companies, and these lists were compared with the lists of deaths from cholera at the General Register Office. There are several reasons, however, why an inquiry thus conducted could only supply an approximation to the truth, and could bear no comparison, in point of accuracy, with a personal inquiry made on the spot, at the time of the epidemic.

1. The inquiry of the Board of Health is into the water supply of the house where the death took place, and not, like the previous inquiry, into that in which the fatal attack occurred; but many persons attacked with cholera in houses supplied by the Southwark and Vauxhall Company were removed to workhouses supplied by the Lambeth Company, whilst hardly any persons were attacked in houses supplied by the latter company, and then removed to a workhouse having the opposite supply, as I know from my personal inquiries in the first seven weeks of the epidemic.

2. Throughout the greater part of Lambeth, Newington, and the Borough, the houses were either without numbers, or numbered very irregularly; and the numbers were liable to frequent change, especially where new houses are constantly being added: therefore, numerous errors were liable to be made in comparing the lists. There were often two or three houses of the same number in the same street; thus it happened that, in the first fatal case I inquired about, the death did not occur at the first No. 6 I called at, but at the No. 6 down the other side of the way; and the water supplies of the two houses were different. Now this is particularly important; for, as the deaths were six times as numerous in the houses supplied with impure water as in those with the better supply, the result would be that, out of every six mistakes, five would transfer a death from the former houses to the latter, and only one would transfer a death from the latter houses to the former.

3. It so happened that the lists supplied by the Lambeth Water Company (that with the purer water) are so arranged and explained that every place might be made out, unless when the above mentioned difficulty about numbers occurs; but the lists supplied by the Southwark and Vauxhall Company are made out in such a manner as to be of only very partial service. They have a kind of alphabetical arrangement, but it is of no use. For instance, to put down such names as Albert Terrace and Providence Place, with no other information than that they are somewhere in a district which extends over ten or fifteen square miles, is to give very little information. Consequently, whilst all the deaths occurring in houses supplied by the Lambeth Company could be identified in the list, and others attributed to these houses from the sources of error above mentioned, it would necessarily happen that a great number of deaths occurring in houses supplied by the Southwark and Vauxhall Company could not be identified; and, in the Report of the Board of Health on Cholera, as affected by Impure Water, as many as 1,436 deaths, in the epidemic of 1853-54, are returned as occurring in houses supplied from unknown sources, although there were comparatively few houses which were not supplied by one company or the other.

The deaths in the epidemic of 1853 are included, with those of 1854, in the Report by the Board of Health which is quoted in the Journal of the Association; but this circumstance could not much affect the result, and certainly not in the direction in which it deviates from the original inquiry; for in 1853 there [864/865] was but a small number of deaths, especially in the districts to which the water supply of the Lambeth Company extends.

For the various reasons stated above, we may conclude that the supplemental inquiry of the General Board of Health into the influence of water supply on cholera is of some value, and corroborates the original investigation, but ought by no means to be quoted as an exact exposition of facts, or be allowed to set aside the previous inquiry.

bottom of book image