Apert: The Laws of Naudin-Mendel 



495 



It is obvious that, in a general way, 

 the proportions found in the geneal- 

 ogies are quite close to the proportions 

 calculated theoretically. It is noticeable 

 that, in general, they are somewhat 

 higher. It is not unnatural to conclude 

 that some factor has come into play, 

 which adds its influence to the heredi- 

 tary factor ruled by Mendel's Law; 

 should we invoke, beside the heredi- 

 tary factor, a factor of innateness, as 

 Lucas does? Perhaps, because it is 

 certainly necessary that albinism should 

 have appeared on some occasion, for 

 the first time; perhaps this factor still 

 exists, to add its influence to the factor 

 of heredity. But to me another explana- 

 tion seems more ])lausible. Note that 

 the discordance only exists in the small 

 phratries; it disappears in the phratries 

 above eight subjects. I think it is be- 

 cause these researches do not repre- 

 sent the population in its entirety. I 

 mean that the authors have gathered 

 their material by collecting all observa- 

 tions published on the subject of albin- 

 ism: now, it is very certain that not all 

 cases of albinism are published, but 

 only those which present some inter- 

 esting particulars; among the interest- 

 ing details which would lead to pub- 

 lication is the coexistence of several 

 albinos in the same phratry, while an 

 observer would easily pass without 

 publication an observation relative to a 

 single, isolated albino, unless he be- 

 longed to a very large phratry, in which 

 case his isolation would itself be an 

 interesting fact. In small families, the 

 facts published by preference are those 

 in which albinism appears relatively 

 frequent; and it is in this manner that 

 I should explain how the proportion of 

 albinos found in small families is some- 



what larger than the calculated expec- 

 tation. It is probable that if the statis- 

 tics could include all families in which 

 albinism exists, the discrepancy — small, 

 as it is — between observation and theory 

 would be much diminished. We would 

 then have only slight oscillations above 

 or below the theoretical probability, as 

 we have at present in the families of 

 nine, 10, 11, 12 and 13 children. 



To ■sura up: the figures brought out 

 of the documents are in very close 

 agreement with those required by the 

 Mendelian law, or present only diver- 

 gences explicable by the conditions 

 under which the documents were com- 

 piled; and all these divergences are in 

 the direction which the conditions lead 

 us to expect. 



We can even extend this agreement 

 farther still. It is easy to calculate, 

 from the mathematical law, the prob- 

 ability of the number of albinos in 

 "albino phratries." Take, for example, 

 the phratries of seven children. Calcula- 

 tion shows one chance in 14,198 that 

 in these seven children there will be 

 seven albinos; 21 chances in 14,198 that 

 there will be six albinos; and 189, 945, 

 2,835, 5,103 and 5,103 chances in 14,198 

 that there will be five, four, three, two 

 and one albinos in seven children. 



Let us compare this calculation with 

 the observation. In the work of Pear- 

 son, Nettleship and Usher, we find 57 

 albino phratries containing seven chil- 

 dren each. One of these contains six 

 albinos, four contain four, 18 contain 

 three, 14 contain two and 19 contain 

 one albino among the seven children. 

 Calculating these figures on the basis 

 of 100, to make comparison eas}^ we 

 get the following table: 



' PHRATRIES OF SEVEN CHILDREN. 



