90 PROCEEDINGS OF THE AMERICAN ACADEMY. 



is exact agreement in but one case (No. 4), the shortage or excess in 

 any one lot was not over 3.75, and this (excess) occurred in but a single 

 case (No. 8). Moreover, as previously pointed out, an exact 3: 1 ratio 

 of pigmented to albino mice is rarely to be exjiected experimentally, and 

 even then only in lots where the number of young is 4 or a multiple 

 of 4. Hence, it is only " in the long run " that the theoretical ratio 

 may be expected to obtain closely. The totals in this case are 1G4 

 pigmented to 51 albino rats (or 23.5 per cent albinos), so that the latter 

 class is but 2.75 individuals, or 1.5 per cent, short of the expected total. 

 The irregularities in respect to the numbers of albinos are very eveidy 

 divided among the eleven lots in which the 3 : 1 ratio was not realized. 

 For in five lots (Nos. 1, 2, 6, 8, 10), there was an excess of albinos rang- 

 ing from 0.25 to 3.75 individuals, with a total of 7.5, as against six lots 

 (Nos. 3, 5, 7, 9, 11, 12), showing a shortage in the number of albinos, 

 ranging from 0.75 to 3 individuals, or a total of 10.25 individuals. 

 Crampe's rats therefore show that complete albinism is inherited among 

 them precisely after the Mendelian principles, though with small num- 

 bers the exact Mentlelian ratio may not be closely approximated. No 

 extended series of crosses in which the ratio of pigmented to albino 

 young is expected to be as 1:1 is given by Crampe. 



In two noteworthy communications von Guaita ('98, : 00) gives his re- 

 sults in breeding black-white dancing mice to albinos of an inbred stock. 

 His experiments were performed manifestly without reference to the 

 Mendelian principles, but nevertheless when the results are examined, 

 they afford additional proof in support of those principles. 



Castle (: OS'^, p. 226) has already shown that as regards the numbers 

 of pigmented and albino young obtained by von Guaita there is almost 

 perfect accord with the total expected on the Mendelian hypothesis that 

 the two characters undergo segregation. The following table (J) is a 

 slight expansion of the one that Castle has constructed, and shows the 

 numbers of albinos expected and those actually obtained, as well as the 

 numbers of pigmented mice. In some cases, the total number of young 

 obtained from a single pair of mice is not sufficient to show conclusively 

 whether or not either of the parents had the albino character recessive, 

 and for these cases an allowance has to be made ; yet their effect on the 

 total result is insignificant. The Mendelian formula applicable to each 

 case is indicated by the figures in parenthesis in the proper column. 



As regards the inheritance of complete albinism, it is seen that from 

 the various crosses, producing in all 309 mice, a total of 77 albinos oc- 

 curred, where the expectation was that there should be from 75 to 79. 



