256 Life and Letters of Francis Galton 



I am quite sure that my way of working on the 2 forefingers is the best for getting at the 

 relations between the various patterns, and I have already learnt much that is new, but I 

 shrink from more work on my present material. Finger patterns seem to me an ideally good 

 subject, not only for heredity work, but for much else of evolutionary interest. If you think 

 the enclosed table of 200 cases full enough, or nearly so, I should take pains to get that 

 number, or double that number, printed off at some school or elsewhere, especially for this 

 inquiry. They would have to be rolled impressions printed in triplicate at least. Such impres- 

 sions are rapidly taken. I easily take 12 of my own fingers carefully in one minute, when all 

 is ready, or in five minutes counting from the time of sitting down to the table with my appa- 

 ratus in my pocket to that of rising with everything cleaned and packed in my pocket again.... 



Very sincerely yours, Francis Galton. 



The weather is becoming cold, not good for travel further. 



[Postcard] University College, London. March 6, 1903. 



I have just had time to work out your correlation table. I find : 



Mean, Left forefinger - 244 mean ridge interval. 



S.D. „ „ -1047 „ 



Mean, Right „ -229 „ 



S.D. „ „ -1048 „ 



Correlation = -8203. 



The correlation of the distal phalanges of the R. and L. forefingers as given by Lewenz and 

 Whiteley in Biometrika is -79, i.e. within probable error of your result. Karl Pearson. 



I wrote further to Galton asking for information upon the "interspace" 

 and upon the want of continuity due to the Arches being treated as of zero 

 "interspace." One of the main difficulties in his restriction of the data to 

 the two forefingers was that a rare type of print that appears on one fore- 

 finger may not appear on the other forefinger but on some* other finger of 

 that hand, and experience seems to show that the prints of all ten digits 

 must be taken into consideration when judging the resemblance of relatives 

 by means of finger-prints. I received the following illustrative letter from 

 Capri : 



March 16, 1903. 



My dear Karl Pearson, I have at last got your long letter of March 5 and enclosures 

 at Capri, where we have been 9 days. After a very little more touring we turn homewards. 

 The loss of 20 subscribers is bad, and so is the attitude of both biologists and mathematicians 

 to Biometrika, but the second year of a new venture is always the most trying time*. The first 

 flush of expectation is over, and the solid merits have not had time to assert themselves. It 

 seems to me to want some cheery writing in good reviews to show in an intelligible form a few 

 definite blunders into which biologists have fallen for want of biometric methods. I expect 

 craniology would furnish topics. I recollect once that kindest of men, Sir W 1 " Flower, being 

 on the verge of wrath because I pointed out the insufficiency of evidence drawn from the mean 

 values of a few skulls of some savage race (I forget which) in determining the race to which 

 a particular unknown skull belonged. Craniological literature would contain, I should think, 

 many rash statements which could be assailed triumphantly by a facile writer and sharp critic. 

 Dear old Beddoe is the most rambling of thinkers and writers as well as one of the most 

 industrious of workers. I am not surprised to hear that his paper is far below the occasion, 

 wrong in its criticism and wrong even in its arithmetic and generally slipshod. The photo- 



* [Those were anxious days for the Editors of Biometrika as they watched the slipping away 

 of their funds. Nowadays several parts of those early years have been reprinted, and complete 

 series sell at very high prices !] 



