HERKDITAKY TKANSMISSION OK GENOTYPICAI. DEAF-MUTISM 37 



typical deaf-mutes. The consequence is that we get more than 25 % 

 affected persons when making such comparisons. This state of 

 things 1 noticed in regard to the recessive mj'oclonus-epilepsy in 

 the large family of Lister which I have examined. 



In 1912 I discussed this very state of affairs with the well- 

 known statistician Weinberg of Stuttgart, and he worked out a 

 method of calculation on the basis of it and indicated a formula 

 which I applied for the correction of this source of errors (see my 

 work: Mediz.-Biologische Familienforschungen, Jena 1913, pp. 

 452 ff.). 



The myoclonus-epileps3% as we now know, is a disease transmit- 

 ted as a recessive homogeneous character. Upon combining my own 

 material with foreign I found that the disease occurred in 19 families 

 with 37 affected out of 113 descendants. That makes 32,7 %, but after 

 the correction by Weinbergs formula the percentage is 25,i. A 

 better agreement could scarcely be imagined. 



If Bergh had read this statement of mine, he need not have 

 groped in the dark about this matter, as he now does in his treatise. 

 He writes: »After this it will be my task to examine to what extent 

 my material shows an agreement with Mendels law. There occur 

 in my material sixteen families among whose numerous offspring a 

 manifold appearance of congenital deaf-mutism occurs. So there is 

 reason to suppose that the deaf-mutism of those children was consti- 

 tutional. 



If we assume that their parents were heterozygotes, we should 

 expect 25 % of the children to have been born deaf, and 75 % of 

 them to be in full possession of their senses. This is, however, not 

 in agreement with the actual state of affairs. In the families men- 

 tioned there were born 105 children in all, and of these 38 = 36, i9 % 

 were born deaf, and 67 = 63,8i % were normal. The explanation 

 of this considerable discrepancy can only be, either that the number 

 of normal children has not been correctly stated to me, a possibility 

 which I am not at all inclined to entertain, or else that the absolute 

 number of individuals was too small, and that this factor affected 

 the relative numbers.» 



As already mentioned Bergh's conclusions are too hasty. If 

 we apply Weinberg's method we find that Berghs numbers agree 

 very well with the theoretical 25 % to be expected according to 

 Mendelian rules. Thus they afford a very decided support to my view. 



Bergh has further found that in 5 marriages between constitu- 



