The Process of Heredity 543 



Experimental Data 



The facts that the two species show such marked physiological 

 differences and that these differences exhibit themselves when the 

 embryos or adults are subjected to sub-optimal conditions, led to 

 experiments that would afford a comparison of the resistances of 

 the pure and hybrid strains at the same stage of advancement. 

 This type of experiment requires rather large numbers of specimens 

 and a considerable amount of repetition, on account of the high 

 degree of individual variation, especially among the hybrid strains. 



I Resistance to Lack of Oxygen 



In 1905 considerable time and effort were expended in the en- 

 deavor to determine the comparative resistances of the pure and 

 hybrid strains to lack of oxygen. Engelmann gas chambers, boiled 

 sea-water, and a continuous stream of purified hydrogen were used. 

 At that time it was not known that the resistance to lack of oxy- 

 gen varied with the degree of development, hence a mass of very 

 contradictory data was obtained. The following summer the 

 experiment was repeated several times with embryos of the same 

 age and degree of advancement. The results were quite different 

 from what one would naturally expect. The H hybrids proved 

 to be the most resistant, the H pures next, the M pures next, and 

 the M hybrids the least resistant. The two hybrid strains were 

 the extremes while the two pure strains were the means. 



These experiments were so tedious, necessitating watching for 

 nearly twenty-four hours, continuously, or at frequent intervals, 

 that another method of testing the comparative resistance of pure 

 and hybrid strains was tried. This method consisted of putting 

 the embryos in gas chambers divided into four compartments of 

 equal size, filling the chamber with sea-water charged with carbon 

 dioxide, sealing up the chamber and timing the cessation of heart- 

 beat in all specimens. Of course only small numbers could be 

 used, because it was necessary to keep a very close watch over all 

 specimens in order to get even approximately accurate death rates. 



Experiment i, June 26, 1907- The M hybrids showed the 



