Correnpon(leti(-r trith Charles Daririn 179 



Now you may perlmpv like t<> hour a few addiliunal |>articuliini aliout myself. I cannot 

 remeiiiltier thu time whon I had not a [Muuiion for colloctiiij;, — fintt ftoalii, frankii, then niitif raU, 

 shMn etc. Ah far as I am oonftciouM, thi! one compulHory fx*-rc:iHO <lurinf( my school life which 

 improvt^l my iiit<)llect wtiH <luin>( Kuclid, and this wax parity voluntary. 



At Kdinlxirgh I do not think the loc-turnH weru uf any service to mo; hut I profited •« & 

 naturalLst by ohju'rvin^ for myKcIf marine animals. 



At C^amliridjjti j{ettin>{ up Pali-y's Kvi(h'ncfH and Mural Phil, thoroughly well aa I did, I felt 

 wa.s an mlmirnlilu Iraininii;, and «verytliing ela4; l>ONh. 



Aly (<<lucittion really Ih'^hii on Ixjjinl the "Heagle." 



I niu.st mid that my son Frank said he could safely give as my character, "sober, honest 

 and industrious." 



And now I want to ask you a question: if I had 50 men of 2 different nations, and for 

 some reiuson eoulil not measure all, if I pickt'd out the 10 tallest of ejjch nation, would their 

 mean heights probably give an approximate mean between all ')0 of each nation 1 



I hope you will get full answers Ui your <iueries, as I dare .say the results will Iw interesting. 



My dear (Walton, Yours sincerely, Ch. Darwin. 



42, Rutland Gate, S.W. May 30/73. 



My dear Darwin, I am truly oblifjed by the Schedule. A few others are sent, many are 

 promised and I have much hopes of useful statistical result in many ways. All I have thus 

 far got confirms the belief that the families will be on the average very small. As for what 

 the usual ixlucation will have l)een, I cannot yet guess. 



In reply to your query about the 50, there seems — or it may be that I ani stupid — that a 

 word is omitted, displaced or someliow wrong, because the .sen-se is not clear and I don't know 



how to interpret the meaning of the phrase " would their mean heights pn)ljably give an 



approximate mean between all 50 of each nation," but the following will probably include what 

 you want. 



If nothinj; else could lie assumed al)out the two nations than that the 10 tallest out of 50 

 taken at liiipliU7.ard from A had a mean height of a', and those from /i of ^, it would bo im- 

 possible therefrom to dwluce either: — 



(1) o and /3, the respective mean heights of the 50 A and the 50 /i or 



(2) the nitio of o to /3. 



But if you grouped the 10 tallest in either case according to their height-s, that is, so many 

 between 5' 10" and 5' 11", so many between 5' 11" and 6' 0" etc., it would be possible by com- 

 paring the run of these numbers with those of an ordinary Table of the Law of Error, to 

 e.stimute appro-ximately lx>th ( I ) and (2). 



10 is too small a numU'r to be serviceable I should fear in this wa}';^l(X) ought to give 

 excellent results; in any case the degree of regularity with which the numbers hapjtened to run 

 would Ix" the measure of the probability of the accuracy of the results. 



If you have any case you want worktsl out aud would send me the figures I will gladly do 

 it Ever sincerely yours, Francis Galton. 



For the year 1874 there are no letters. Darwin was ill in Septeml>er 1 873, 

 Mrs Tertiu.s Galton (Violett^i Diuwin) died in February 1874, Mra Francis 

 Ualton was very ill in 8e})teinl)er and (Jalton hiniselt" at Christinas with 

 " irregular gout and influenza." Darwin's eldest son George (later Professor 

 Sir tJeorge Darwin) takes up the correspondence. 



We return to-morrow to 42, Rutland Oatk. Aor 16/74. 



My dear Georok, Thank you kindly for your letter. My wife was alarmingly ill with 

 a sudden vomiting of arterial bliKsl, rejH'ated during the night but fortunately never afterwanls 

 rtHiurring. .She wius extremely weakent-d and unalije to move out of UhI for day.s, or out of the 

 liouse where we were staying for we<>k8, but she has steadily mendetl and now 9 weeks have 



83 S 



