Varietäten, Descendenz, Hybriden. 357 



nes), may show another one in a low degree (e. g., the measure of 

 the diameter); in other strains the reverse combination may be 

 found. Thus the positive correlation of all characters in the popu- 

 lations is only the expression of an average condition, which may 

 not hold when particular strains are compared. The combinations 

 of hereditary characters distinctive of particular strains therefore 

 can not be accounted for as due merely to the difference in some 

 underlying character (as for example the size of the body.) 



When a Single family is studied by itself (all the individuals 

 being descended by fission from one original parent) a considera- 

 ble degree of correlation between parent and offspring is still found 

 to hold for most characters. Thus within the Single family the 

 offspring resemble their immediate parents more than they do more 

 distant members of the family. In some characters (e. g. the num- 

 ber of teeth, often also the diameter), the correlation of parent and 

 offspring within the Single family is very high (at times 0,5 or 

 more). In some cases this high correlation is not due to inheritance 

 of parental diversities, but to a mere steady increase in size from 

 generation. But in most of the families neither this nor any similar 

 explanation can be given; the correlation is due to the inheritance 

 of parental diversities. 



Selection for diversities within the Single family was carried on 

 in three large families with reference to the number of spines, to 

 the size of the animal and to the length of spines. In all these 

 respects selection was effective. With respect to the number of 

 spines, selection acts slowly; with respect to the other two charac- 

 ters its action in producing diverse Stocks is much more rapid. 

 Number of spines and size were found to be correlated, so that it 

 is not clear that these two characters are acted upon independently 

 in selection. But length of spines is not correlated with the other 

 two, so that hereditary diversities in length of spine are brought 

 about independently of changes in size and number of spines. 



In general the investigations show that in Difflugia Corona a 

 Population consists of many hereditarily diverse Stocks; and that a 

 Single stock, derived by fission from a Single progenitor, gradually 

 differentiates into such hereditarily diverse Stocks; so that by selec- 

 tion marked results are produced. M. J. Sirks (Wageningen). 



Jennings, H. S., The numerical results of diverse Sy- 

 stems of breeding. (Genetics. I. p. 53 — 89. 1916.) 



The paper gives, as summarized by the writer, formulae for 

 finding in any generation the results of continued breeding by a 

 given System, with respect to a Single pair of alternative charac- 

 ters. Sex-linked characters and typical characters are dealt with 

 separately. Formulae are given for tbe results of: random mating; 

 assortative mating; selection of dominants; seif ferülization; inbreeding 

 of brother by sister and of parent by offspring (several Systems). In 

 each case the diverse results obtained by beginning with different 

 parental combinations are given. 



It is shown that the results in successive generations form frac- 

 tions such as may be obtained by compounding in various ways 

 several well known arithmetical series. The first twenty terms are 

 given of fourteen such series; the formulae given show how the 

 terms of these series are to be compounded in order to give the 



