168 CHARLES R. STOCKARD AND GEORGE N. PAPANICOLAOU 



of Fi individuals are worse than the records of the sum of all 

 alcoholic generations, as is seen on comparing the third column 

 with the second. The later generations being further and 

 further removed from the treatment and having less and less 

 modified germ plasm on account of the constant introduction of 

 normal stock are progressively^ improved until finally the F4 gen- 

 eration has its modified germ plasm diluted to such a degree that 

 its record is on par with the control. 



The ancestors of these late-generation animals were also suc- 

 cessively selected from the least affected of the alcoholic stock, 

 being those animals capable of survival and reproduction, while 

 the most highly affected died or were sterile and incapable of 

 reproduction. We assume the probability that the more nearly 

 normal animals with stronger bodies also carry germ cells that 

 are less affected than those in the more degenerate individuals. 

 Most of the grossly defecti\x individuals which reach maturity 

 are sterile as evidence in this direction. Thus individual 

 selection being in this case a selection of germ plasm as well 

 as soma, helps materially to improve the quality of the later 

 generations. 



8. A COMPARISON OF ANIMALS FROM DIRECTLY TREATED FATHERS 



AND FATHERS OF ALCOHOLIC STOCK WITH ANIMALS FROM 



DIRECTLY TREATED MOTHERS AND MOTHERS OF 



ALCOHOLIC STOCK AND WITH OTHERS FROM 



BOTH PARENTS OF ALCOHOLIC STOCK 



Are the general conditions induced by directly treating the 

 father with alcohol the same as those resulting from treating the 

 mother, and are they equal in extent? Do fathers of alcoholic 

 ancestry beget offspring of better or worse quality than off- 

 spring produced by mothers of similar alcoholic ancestry? Or 

 are the effects of the alcohol treatment on the germ cells, which 

 is expressed through several generations, carried with equal 

 degree by both the alcoholic father and the alcoholic mother? 

 We shall attempt in this and the following section to supply 

 data which may serve to partially, at least, satisfy these queries 

 as well as furnish an analysis of several other more detailed 

 propositions. 



