138 ARTIFICIAL PARTHENOGENESIS AND FERTILIZATION 



In order to make the above discrepancies comprehensible, 

 we must remember how many variations of constitution come 

 into play in this experiment. We saw first that the effect of 

 the acid increases with the number of carbon atoms, secondly 

 that the entry of an HO-group has the opposite effect, and 

 thirdly that the " linear" coupling of the carbon atoms is more 

 effective than the "branched' 1 (oxybutyric and oxyisobutyric 

 acids). All these and other conditions of constitution prob- 

 ably come under consideration in the explanation of this appar- 

 ently irregular acid effect of the last table. The number of 

 acids investigated is not sufficient for a detailed analysis. 



It is even truer of the mineral than of the dibasic organic 

 acids that they are able to cause membrane formation only in 

 the eggs of some, but not all females. With sulphuric acid in 

 particular I have never yet been able to evoke membrane 

 formation in the sea-urchin egg; 1 I have done so occasionally 

 with HN0 3 and HC1, but not always. The most favorable 

 results that I have obtained with these two acids are summed 

 up in Table XXIII. 



TABLE XXIII 



Usually, however, the results are more of the kind given 

 for comparison in Table XXIV. 



1 It seems to be a general experience that sulphates diffuse less easily into 

 living cells than chlorides. 



