6o4 



NATURE 



\_April 2^, 1889 



owing to the difference in the mean stature of the sexes. 

 Mr. Galton gets over this difficulty by multiplying all his 

 female data by a proper constant— pushing them up, so 

 to speak, to the male standard. Upon this understand- 

 ing, suppose that the mean height of the father and the 

 " transmuted " mother — the stature of the " mid-parent " 

 in Mr. Galton's phraseology — differs from the mean 

 height of the general population by say three inches. 

 Then the mean height of the sons and " transmuted," or 

 pushed up, daughters, is most likely to be not three inches, 

 but two inches. The constant of " regression " is deter- 

 mined with equal precision for other relationships. A 

 general idea is obtained of the extent to which the pecu- 

 harities of an individual are likely to be shared by his 

 kith and kin. 



It is not easy in a few words, or perhaps in any num- 

 ber of words unaccompanied by symbols, to do justice 

 to the cogency and precision of this anthropometrical 

 reasoning. The manipulation to which Mr. Galton's 

 materials have been subjected by one of his mathe- 

 matical coadjutors, Mr. Hamilton Dickson, fully attests 

 their consistency and strength. Some additional corro- 

 boration may be afforded by the following considerations. 

 The probable error or dispersion for the statures of adult 

 men, which Mr. Galton has extracted from the family 

 records submitted to him, is identical with, or differs only 

 by a fraction of an inch from, the constant furnished by 

 many other sets of measurements. The value here ob- 

 tained for this constant is 1 7 inch. The same value is 

 obtainable from the measurements made by Mr. Galton 

 for the British Association. The same value has been 

 obtained by Signor Perozzo for the whole of Italy, and 

 for each of its provinces. The agreement of observa- 

 tions made under such different circumstances is calcu- 

 lated to give us confidence in the higher theory of 

 anthropometry. The result which has been thus verified 

 may be used to confirm Mr. Galton's reasoning at several 

 points. Any scruples which he may suggest as to the 

 discrepancy in the values of mean stature determined 

 from his different records are removed by a considera- 

 tion of the error or diversity to be expected among these 

 results. Again, consider those tables in which Mr. Galton 

 compares the heights of a number of persons with the 

 mean height of their children or brothers, in which it is 

 shown, for instance, that men of the height 71 '5 inches 

 have brothers averaging 70-2. That all the entries point 

 in the same direction — namely, that of " regression " — is 

 in itself adequate evidence of that fact. But not only 

 are the faggots strong in their union, but also each indi- 

 vidually is possessed of considerable strength. Thus, 

 the discrepancy which we have just noticed between the 

 height of a man and the mean height of his brethren, 

 namely, i'3 inch, is founded on eighty-eight instances. 

 The chances against this degree of divergence oc- 

 curring by accident are some hundreds to one. The 

 odds that the appearance of law which the tables present 

 is not accidental are immensely increased by this con- 

 sideration. We should be curious to know whether Mr. 

 Galton's experiments on the " regression " of sweet peas 

 would admit of this sort of corroboration. 



The human stature is a subject particularly well 

 adapted to Mr. Galton's exact methods of measure- 

 ment. Length admits of more exact gradation than 



the so-called secondary qualities. To arrange in a re- 

 gular scale the colours of eyes which are variously de- 

 scribed as dark blue, blue-green, hazel, and so forth, is 

 a delicate task. How far Mr. Galton has triumphed 

 over this imperfection of his data, it must be for special- 

 ists to decide. The student of probabilities cannot doubt 

 that the correspondences between his observations and 

 his calculations are indicative of a real law. The co. 

 herence of the table in which he compares fact and 

 theory as to the number of light-eye-coloured children 

 born to parents of various eye-colours cannot possibly 

 be accidental. Ill-adapted as eye-colour may be to exact 

 measurement, it is a more satisfactory quality to deal 

 with than " the artistic faculty." Can we suppose that 

 the compilers of the different family records which Mr. 

 Galton has analyzed have employed at all similar 

 standards, when they applied the epithets " artistic " and 

 " non-artistic " to their relations ? Our misgivings increase 

 when we go on to apply the calculus to the returns as to 

 disease which are obtained from the family records. To 

 arrange parents and children in a graduated scale of 

 " consumptivity," upon the testimony of unprofessional 

 relatives, seems precarious. The author himself abandons 

 the use of the more delicate methods when he goes on to 

 consider " good and bad temper." He has not, however^ 

 shrunk from dividing into five degrees or classes some 

 sixty shades of temper ranging from "amiable" and 

 "buoyant" to "surly," "uncertain," "vicious," and 

 " vindictive." 



We ascend into a region of hypothesis when we specu- 

 late on the causes of the phenomena which have been 

 evidenced. The attention of biologists should be called 

 to Mr. Galton's views on " particulate inheritance, 

 " latent characteristics," and the stability of organic 

 forms. The conceptions which he has formed as to the 

 processes of heredity are placed by him in a variety of 

 lights, and illustrated by many happy analogies. " Ap- 

 propriate and clear conceptions," it has been well said, are 

 essential conditions of science. Mr. Galton has done 

 much to make his abstract ideas clear, but are they also 

 appropriate ? This is a question upon which, perhaps, 

 only a few specialists are competent to advance an im- 

 portant opinion ; and their authority is liable to be im- 

 paired by the prejudices incident to an exclusive line of 

 research. We shall be slow to accept adverse criticism 

 from any whose studies may not have qualified them to 

 appreciate the support which Mr. Galton's theories receive 

 from his masterly use of the calculus of probabilities. 



F. Y. E. 



NATURE'S HYGIENE. 

 Nature's Hygiene: A Systematic Manual of Natural 

 Hygiene, containing a Detailed Account of the Chemistry 

 and Hygiene of Eucalyptus, Pine, and Camphor Forests, 

 and Industries connected therewith. By C. T. Kingsett, 

 F.I.C., F.C.S. Third Edition. (London: Bailliere, 

 Tindall, and Cox, 1888.) 



TH IS book aims at being a systematic manual of natura 

 hygiene. The introductory chapters deal in a popular 

 manner with chemical principles and chemical changes, 

 leading up eventually to questions affecting the chemistry 

 and hygiene of the atmosphere, of water, of sewage, and of 



