a single line of Forel's statement. It is only reasonable to suppose that Dr Saleeby has never seen Bezzola's "well-known work" which he finds "most conveniently quoted from Forel." Had he done so he would have known that (i) the data were not taken from the Swiss census of 1900, (ii) there were nearer eight than nine thousand imbeciles, (iii) there are not two acute annual maximum periods for the conception of idiots, (iv) there is no evidence for the conception of idiots at the time of the vintage being enormous and almost nil at other periods in either the wine-growing districts or any others*, (v) the two periods of maximum conception of idiots do not coincide with the times of the year when conception is at a minimum in the rest of the population. In fact no statement made by Forel in the extract cited by Dr Saleeby is correct t. But we have not done with Dr Saleeby yet! In his paper in the British Journal of Inebriety, Vol. v. p. 59, he tells us: "We have Bezzola's inquiry showing that in Switzerland most idiots are conceived at the time of the vintage." Will the reader again fix his attention on our diagram and note how the excess of three idiots, lying in itself well within the limits of random sampling (see the point ẞ and carry the eye to the bottom of our diagram!), has now been magnified into "most idiots"?

Dr Basil Price-honorary treasurer of a society which professes to study inebriety-finds the memoir of the present writers "effectively criticized" by an author who can make such statements as we have just cited above. He himself contributes to the humour of the situation the following paragraph (British Journal of Inebriety, Vol. VIII. p. 73):

"In certain wine-growing districts of Austria it has been shown that the majority of imbeciles are conceived during the periods when most drinking takes place."

Here we have a perfectly definite statement as to a majority of imbeciles being conceived, although it is delightfully vague as to what the periods in question are. This paragraph has been taken as confirming Bezzola's results (?) as to the vintage. On what is it based? Dr Basil Price merely cites Leppmann in Senator and Kaminer without reference to a page. The only passage we can find bearing on the subject in Leppmann's paper occurs on page 1092, and runs: "At the discussion on this interesting communication (i.e. Bezzola's) at the Vienna Congress against alcoholism, a medical man said that the teachers in wine-growing districts of Lower Austria know that a material of very bad scholars in any one year denotes a good vintage years previously."

This statement made by a nameless man‡ as to the opinion of teachers as to

* The argument as to the vintage would be almost exactly parallelled if September were shown to be a month with an excess of imbecile births in Kent, and it was then attributed to beer, because it was the month of the hop-harvest. Are the "Weinleser" supposed to consume the grapes or the must?

+ Yet Dr Saleeby writes: "The well-known research of Bezzola...was based upon an enormous number of cases derived from official statistics. The evidence has been studied and accepted by such a careful critic as Professor Forel, and the precise data for which Dr Ryle asks [evidence of the results of conception during acute alcoholism], as if we had them not, are the very data the careful analysis of which led Bezzola to his conclusion" (National Temperance Quarterly, Sept. 1910, p. 170).

It is actually quoted again as scientific evidence by Sir Victor Horsley (Alcohol and the Human Body, p. 326).

bad scholars for the whole of one year-an absolutely worthless statement for any scientific purpose, with not a number in it nor a measure of "bad" and "good' is directly modified by Dr Price into an assertion that it has been shown, i.e. demonstrated, that a majority of imbeciles are conceived in Austrian wine-growing districts during the periods when most drinking takes place! Can any better proof be given that the foremost medical champions of degeneracy as a product of alcohol in the parent at the time of conception are entirely untrustworthy as critics and as men of science? But there is a more serious aspect to the whole matter. The British Journal of Inebriety is the organ of the Society for the Study of Inebriety, and on the first page of this Journal is a list of the vice-presidents of this society. It includes among other well-known names those of Sir Clifford Allbutt, Professor William Osler, Dr F. W. Mott and Dr R. Welsh Branthwaite, men who certainly weigh their arguments before they speak. Have these leaders of medical thought any idea of the gross absurdities to which they, unconsciously perhaps, but none the less effectively, are giving currency by the appearance of their names in this Journal? We had no desire whatever to examine the literature of this subject before starting our own investigation; we determined to come unprejudiced to an examination of the problem on unbiased data. This we accomplished, and we published our results indicating that the consumption of alcohol in the parent did not produce any marked mental or physical degeneracy in the offspring of school age. We have been assailed by a perfect army of critics-largely furnished with old-fashioned blunderbusses-who apparently confuse the whole alcohol problem with the problem of the influence of parental drinking on the mentality and physique of the offspring. They sympathise with Laitinen when he asserts that one glass of beer a day exercises a degenerative influence on the offspring or with Bezzola when he anticipates that: "Man wird vielleicht einmal zur Einsicht gelangen, dass jeder Tropfen Alkohol beim Erzeuger einen Tropfen Dummheit bei Erzeugten bedeutet ' (Internationale Monatsschrift zur Bekämpfung der Trinksitten, 1901, p. 183).

These extreme views may or may not be ultimately demonstrated to be correct. We are quite certain that at present there is no definite evidence at all in their favour. We are now also fairly convinced that evidence collected by such critics as we have referred to will never be treated by any scientific method that we could trust, and that we should further have to ask overwhelming proof of the unbiased nature of its origin. We believe we have sufficiently illustrated in our sample of three stock temperance medical memoirs how wholly lacking these writers are in critical instinct; any processes or any data are sufficient, if they can be made to support a preconceived opinion*. Such authors can on one page speak of the "thorough

* Thus Sir Victor Horsley tells us that we have wholly neglected "Galton's well-known law that the contribution of the parents to the physique and mentality of a child is about one-half of the whole, that of the grandparents one-eighth, and so on. When one parent only is considered (see memoir) this gives us only one-fourth of the total hereditary forces at work, and consequently no conclusions whatever ought to have been drawn" (National Temperance Quarterly, Sept. 1910, p. 181). Now Sir Victor Horsley is one of the persons who assert that alcohol produces a toxic effect on the germ and so on the offspring. What has this to do with Galton's law, which, whether true or not, was stated by its author as

G. M.

4

investigation" of Dr MacNicholl, and the next page state that the absence of any inquiry as to the "habits of the parents previous to and during conception of the children" quite vitiates any value in a report (Dr Basil Price, N. T. Q. loc. cit., pp. 176-7). Such authors can in one breath tell us that although we cannot for obvious reasons find out the exact condition of the parents at the conception it is possible to find out "their habits of life," and yet in the next assert that the "habits of life" are not sufficient but that we require knowledge of three or four generations of the stock! In short everywhere logical confusion, chaos and misstatement.

One remarkable point remains to be noticed. In April 1901 a Report was presented to the "Society for the Study of Inebriety" by their Committee on Heredity consisting of nine medical men. This Committee state that:

"They are aware of and have devoted full consideration to the widespread belief that parental indulgence tends to render the offspring more innately prone than they otherwise would have been to excessive indulgence, but they can only reiterate their conviction that the existing evidence on the subject does not at present warrant such a conclusion" (Art. XIV). And again:

"In particular there is no evidence that characters acquired by the parent through indulgence in drink are inherited by the children subsequently born. The committee are aware that it is possible that the mental and physical states produced applying to an inherent, not acquired, germinal character? The answer to this question is at once found by examining Sir Victor Horsley's book on Alcohol and the Human Body. He does not know what heredity means in any modern scientific sense. He speaks of "the appalling force of hereditary influence" when he is referring indifferently to the toxic effect of alcohol on the germ, or to an inherent character of the germ-plasm of a given stock. When Dr Crothers says that in reporting 1744 cases of inebriety he found 1080 with a distinct history of heredity, it never appears to occur to Sir Victor that the case may not be the same as that of alcoholised hens' eggs or the puppies of alcoholised dogs or guinea-pigs. Fancy a modern scientist talking about the manner in which "the appalling force of hereditary influence may be mitigated!" The only manner in which you can mitigate heredity is to cease to breed from bad stock, and no toxic effect of either alcohol or syphilis has anything to do with true heredity. As for Sir Victor Horsley's statement that no conclusions whatever ought to have been drawn from our memoir "because when one parent only is considered only one-fourth of the total hereditary forces are at work," the reply is obvious that if one-fourth produces zero effect, four times one-fourth will also produce zero effect. The application of the law to such a case, however, is wholly meaningless. Sir Victor Horsley's final conclusion is one of the most sweeping that we have yet come across! "The fact is there is only one way in which this question can be properly studied, and that is by obtaining data from some source which can provide instances of genuinely abstaining families for three or four generations. These should be compared with people in similar circumstances of life, amongst whom it can be proved that drinking habits have prevailed for the same period" (loc. cit. p. 181). Excellent doctrine! We need data covering 75 years or a century of families remaining in the same circumstances. But until that material is forthcoming, why does Sir Victor Horsley write a book on Alcohol and the Human Body, why does he write a chapter on Parental Alcoholism, why does he open it with a quotation that "Hereditary alcoholism is an undeniable fact," and why, above all, does he cite Laitinen, MacNicholl, Crothers and Bunge, who have wholly failed to comply with the "only" way "in which the question can be properly studied" as authorities worthy of consideration in the matter? There is only one answer to these questions. This criticism was not part of his intellectual stock-in-trade, or he would have applied it ab initio to these investigations. He ran up against it as a sort of weapon which might possibly fit the case when he found facts not in accordance with his own preconceived opinions.

in the parent by indulgence in alcohol do affect the child in some way through inheritance; again, they admit as possible, though strictly speaking this is no question of the inheritance of an acquirement, that indulgence may so damage the parental tissues that the germ is ill-nourished and the child is thus affected; yet again they admit as possible that the alcohol circulating in the parents' blood may directly affect the germ and in this manner affect the offspring as by producing degeneracy. But these speculations have not been strongly supported by any evidence tendered to the Committee" (Art. v1).

The words we have italicised hold as truly in 1910 as in 1901, there is no strong support in any evidence yet produced for these speculations. Yet when we come to the same conclusions as the Committee on Heredity of the Society for the Study of Inebriety, on the basis of quite independent and unbiased material, the Journal of that Society has not invective enough at its disposal to describe our "mischievous and indefensible contribution to the printed folly of nations." To such rhetoric we can but reply in the words of the old song:

Pray, Goody, please to moderate the rancour of your tongue,
Why flash those sparks of fury from your eyes?
Remember when the judgment's weak, the prejudice is strong.

(5) Demme's Contribution to the Subject. We will now pass to a last illustration of the manner in which our medical critics deal with statistics and the amount of care they give to their critical examination. We will consider the results of Demme. Horsley quotes Demme from Hodge, Basil Price quotes him from Kirby who again quotes him without any reference to the original at all (British Journal of Inebriety, Vol. VI. p. 166). The tables given by Horsley and Basil Price are not in agreement. With such a method of quoting authorities can we wonder if the title" scientific" be denied to papers published by these temperance writers? It almost compels one to believe that they have not themselves studied, and do not want others to study, the originals from which their data profess to be drawn. We shall content ourselves by criticising the data of Demme as they are actually cited without comment or warning by Sir Victor Horsley and Dr Basil Price.

The following is the record of the children of 10 drunkards and 10 sober pairs of parents prepared by Professor Demme and provided by Dr Basil Price:

Now either these 10 cases of sobriety and 10 cases of inebriety are typical of the whole population of sober and drunken parents or they are not. If they are selected cases then Sir Victor Horsley and Dr Basil Price have no logic whatever when they cite them as exhibiting the influence of alcohol in the parent. We must start therefore with the assumption that they represent the average results of sobriety and inebriety*. Let us see whither it leads us. The percentage of drinking parents in Switzerland must be at least 30%. This is, as the American and English returns show, quite an average percentage, and Professor Demme himself tells us that 10% of the population of Berne die of alcoholism. (For Switzerland 10.5% of men from 20 to 40 years of age and 15.5% of men from 40 to 60 years are said directly or indirectly to have died of drink in the 15 greater towns between 1891 and 1899 †.) We can now reconstruct our total population of which our 10 sober people and 10 drunken people are supposed to be random samples. These random samples of parents would produce offspring in the proportions given in the table below. Now is this not another striking instance of the manner in which these gentlemen in the name of science thrust statistics on the untrained public without having the inclination or capacity to weigh them as well as cite them? In the first place we have an infantile death-rate which is wholly out of keeping with any vital statistics with which we are familiar. In the next place the drunkards have practically died out in the next generation, from some percentage like 30 they have dwindled to 2.5°. ! And if it be argued that we have taken too high a percentage of drunkards for Demme's district, then be it noted these results will be not bettered, but rendered far more anomalous! Make your drunkards only 10% of the population-an obvious under-estimate, because Demme tells us that 10% die of alcoholism, Reconstructed Population according to Demme's Samples.

This is a very big assumption, as any statistician would have informed these gentlemen. The probable errors of the results based upon these numbers are enormous. For example, the probable error of 2 deaf-mutes occurring in 57 persons is '94, thus the real number of deaf-mutes might easily have been either 0 or 4; for 5 dwarfs in 57 it is 1-5, so that the number of dwarfs might have been easily either 2 or 8. Beyond this, however, we know that deaf-mutism and dwarfism run in families, so that these probable errors ought not to be based even on 57 units!

† Archiv für Rassen- und Gesellschafts-Biologie, Bd. 1. S. 238. Dr Basil Price (British Journal of Inebriety, Vol. VIII. p. 77) asserts that on the most conservative estimate the alcoholic death-rate in England and Wales is 14; this can hardly denote less than 30%, who drink considerably.