duty in an orderly fashion should suddenly abandon all restraint and begin dividing and subdividing in so useless a manner. The cancer-cell is one that has lost all-or almost all--its functions save that of growth and subdivision, and modern studies seem to show that its growth can continue-given suitable circumstances-to an unlimited extent. The problem of the nature of cancer resolves itself into the question why a cell, or perhaps a group of cells, suddenly takes on the power of incessant and purposeless growth.

And there are no indications that this problem will be solved in the near future. Suggestion after suggestion, some wild and futile, and some based on a clear knowledge of all the known facts and elaborated with great skill, have been brought forward, but each and all have failed to stand the test of further investigation or to satisfy the profession generally. Even the main question, as to whether cancer is or is not a parasitic disease, and due to a living organism, is not yet settled, and respectable authorities maintain each view. The feeling on the subject, which is not perhaps always expressed, appears to alternate at intervals of a few years, and at present the non-parasitic theory is generally accepted by pathologists. There is, however, no definite and conclusive proof against the microbic theory, and signs are not wanting that it may come into favour again. In the past very many organisms, bacteria, yeasts, and protozoan parasites have been thought to be the true cause, and in some cases on what appeared to be very strong evidence. All these have now been definitely exculpated, and there is now no doubt that if cancer is due to a living organism, it is either one that requires very special methods of staining to make it visible, or -and this is more probable-that it is one of the microbes, the existence of which is now proved, that are smaller than the smallest objects that can be demonstrated with the highest powers of the microscope available at present. A very minute bacterium, capable only of living (in the body at least) in the substance of a cell of the patient, and stimulating this cell to continued growth, is a possible conception, and if it occurred it would appear to account for most if not all of the facts known about cancer whether of man or of the lower animals.

The other scientific theories of cancer, those which attribute it to developmental defects, or to chemical changes in the blood or tissues, need not be discussed. In most cases the former are founded on analogies, more or less fanciful, with what takes place during the development of the embryo, a period in which the whole of the energies of the body are directed to the one function of growth, just as in the case with the tumour-cell. Even if we admitted their truth, they would only remove the problem a step back, and the actual exciting cause of the disease would be still to seek. The majority of pathologists at the present day, whilst inclining away from a microbic view, and condemning all theories at present advanced as untenable, join with Newton in saying, 'Hypotheses non fingo,' and concentrate their attention on the discovery of new facts.

With the suggestions which attribute cancer to eating meat, tomatoes, or other vegetables, too much or too little salt; to the presence of sulphurous compounds in the air; to the increasing stress of modern life; to vaccination; to the presence of decomposing animal or vegetable matter, and to a thousand and one other supposed causes-it is not necessary to deal. In all cases the theories are readily disprovable by an examination of the distribution of the disease amongst mankind in different countries and conditions. We are not prepared, however, to deny absolutely that meat-eating in excess may have some effect in increasing the liability to the disease, though its occurrence in vegetarians and herbivorous animals demonstrates that it is not an invariable and necessary cause. One of the most thoughtful and philosophical attempts to find an etiological factor for the disease, i.e. a cause which, acting on a large scale, increases its prevalence, as opposed to the vera causa which stimulates the cell to start its anarchistic career, is that of Roger Williams. According to him cancer is pre-eminently a disease due to increased civilisation, accompanied, as is usually the case, with an increased influx of the rural population to the towns, increased material prosperity, and consequent excessive feeding. Williams accepts the figures which appear to show that cancer is actually increasing rapidly, and attempts to show that this increase is due to the sudden change in

environment which occurs in the passage from poverty to riches, or from a rural to an urban habitat. The figures of the incidence of the disease in different countries tend, on the whole, to bear out the idea that cancer affects chiefly the more civilised communities; yet even here there are difficulties, arising partly from the fact that statistics are better kept in these communities, and that the individuals who compose them are more likely to live on into the cancer age. Williams cites, in proof of his theory, the figures for London in 1903, which show cancer to be most prevalent in Hampstead, Marylebone, and Chelsea, and low in Holborn, Shoreditch, Southwark, Bethnal Green, Finsbury, and Stepney. An even better example is furnished by a comparison of the statistics for Bristol (with its mixed population), its wealthy suburb, Clifton, and its poor one, St Philip :

The comparison between the low death-rate from phthisis, well known to be a disease due to under-feeding and general unhealthy environment, and high death-rate from cancer in the wealthy suburb, with the high phthisis and low cancer death-rate in the poorer locality, is certainly striking, and figures like these, in which the individuals concerned are grouped under conditions approximately similar as regards medical attendance and methods of collecting statistics, must be regarded as of much greater value than those obtained in different countries. The Irish figures bear out the theory very well. In Kerry, where the inhabitants are notoriously underfed, there is much phthisis and little cancer, whereas in Ulster the opposite conditions obtain. On the whole, it is probable that Roger Williams' theory is correct, at least in part, and that a change in environment such as is due to a change from a rural to an urban life, and an abundant diet, especially one rich in meat and other

proteids, must be regarded as being at least one of the factors in causing a prevalence of the disease. There are, however, difficulties. It is hard, on this theory, to account for the low death-rate in Brittany and the high rate in its adjacent Normandy, or the very high figures met with in French towns and the comparatively low ones seen in those in England, and especially America. Thus the death-rate per 10,000 living is, in Boulogne-sur-Mer, 17.5; in Rouen, 16.9; and in Paris, 118; whereas in London it is only 68; in New York, 6-2; and in Washington, 5. Differences in methods of registration hardly appear capable of explaining differences so wide, nor can they easily be harmonised with Williams' theory. Furthermore, it is quite certain that none of the factors on which Roger Williams lays stress are necessary antecedents of cancer, for it occurs in savages as well as amongst civilised folk, in country as well as in towns, and in vegetarians as well as in meat-eaters.

One of the most important questions of the day deals with the supposed increase in cancer. We say 'supposed,' for here, as in so many questions connected with the subject, certainty is not yet attained. One thing is quite clear, and that is, that in nearly all countries in which statistics are kept there is a larger recorded proportion of deaths due to cancer year by year. The following table will show this increase in England and Wales:

Similar results are obtained from a study of statistics in all parts of the world. Thus in Scotland there has been a rise in the death-rate per million living from 416 in the decade 1861-70 to 890 in 1905; the numbers are usually greater than in England. In Ireland (where, however, the value of the statistics is greatly impaired by the extent to which emigration has taken place) the

rate has increased from 270 in 1864 to 793 in 1906. Similar increases are reported from all parts of the world in which statistics are kept, and this both in rural and urban districts.

The interpretation of these crude figures is a problem of extreme difficulty. There can be no doubt whatever that they exaggerate to a very great extent any possible rise that may have occurred, and, in the opinion of some eminent authorities, they do not prove that any increase whatever has occurred. In the first place, a deduction has to be made for the greater number of individuals who survive to the age at which cancer is most prevalent; for the disease is, in all its forms, one of advanced life. This is well shown by the following table, showing the prevalence of the disease at all ages in the city of Paris; the figures give the proportion of deaths from the disease per 100,000 living at that age:

The figures for England and Wales (per million living at the various ages) are:

The two sets of figures (and many others might be quoted) are quite comparable, and show clearly the increasing liability to cancer with increasing age, and they show us that it is quite useless to compare the cancer death-rates of two communities unless the agedistribution of the individuals composing each is known. For example, we should expect to find a low rate in a