BOWMAN LECTURE. LXXI 



(a) Seventeen completed sibships containing total 

 117 (100), affected 55 (48), normal 62 (52). 



(6) Ten similar sibships containing total 58 (100), 

 affected 15 (26), normal 43 (74). 



(c) Three similar sibships, containing total 13, affected 

 1J, normal 2. 



It will be noticed that in these three small sub-groups, 

 where in all cases the parents were normal, only the 

 second (b) fits the Mendelian expectation of Fig. 6 D, 

 where two impure dominants carry the disease as recessive 

 and throw one quarter of their offspring diseased. Both 

 (a) and (c) require dominant to have changed place with 

 recessive in the second generation in order to bring them 

 into the theory at all. (Fig. 6 A and c.) 



The numbers I have just quoted are the outcome of 

 careful examination and the exclusion as far as possible 

 of incomplete examples ; I hope, therefore, that they will 

 not be without interest at the present time. I may say 

 that I Avas quite unprepared for such a near approximation 

 to halves and quarters as are shown by certain of these 

 groups. 



Allusion has been made to the change of dominance 

 supposed to occur in sex-limited disease. I believe there 

 is clinical ground for suspecting that dominance, if we 

 use the term, may sometimes change, or rather may 

 be different, for the same disease in different families 

 irrespective of sex ; and if this be true, the factor causing 

 the alternation of dominance in the sex-limited cases may 

 be, not sex itself, but something else, usually, but not 

 invariably, associated with sex. Retinitis pigmentosa, for 

 instance, appears to be recessive in many families, but in 

 the largest recorded pedigrees it behaves as a dominant, 

 and yet it is the same disease in both instances. If 

 such change can occur at all, we need go only a step 

 further in order to explain the first appearance of a 

 dominant disease. A condition that has for want of 

 meeting with another similar gamete been propagated for 

 generations as recessive in an impure dominant would at 



