J. Henderson Smith 



43 



and the other half to 04 per cent, phenol, and the experiments were 

 carried out simultaneously. The result is seen in Table XX and Fig. 7. 

 The 0-7 per cent, curve shows no sigmoid character and approaches 

 closely the unimolecular type, although in this particular experiment 

 not strictly logarithmic (r. values of k in the table). The 0-4 per cent, 

 curve, on the other hand, is frankly sigmoid. The same spores, then, 

 give a sigmoid or a logarithmic curve according to the strength of 

 phenol employed. The shape of the curve does not depend simply on 

 the character of the spores used, but chiefly on the strength of phenol. 



Survivors 

 TOO 



80 



60 



40 



20 



0-0 1 Minutes 2 3 4 5 



X---X 15 30 45 60 75 



90 



7 

 105 



120 



Fig. 7. Same suspension with x X 0-4 per cent, phenol and 



O O 0'7 per cent, phenol. 



The 0-7 per cent, graph is plotted on a 15 times more open scale than the 0-4 per cent. 



What, then, is the explanation of the logarithmic type of curve? 

 It is evident that the explanation given for the sigmoid 0-4 per cent, 

 curve will not serve here, at least not without modification. The 0*4 per 

 cent, curve, when transposed into a frequency curve in the manner 

 described, gives an approximation to a normal curve {v. Fig. 8), and 

 the explanation derives its convincing cogency from the fact that the 

 derived frequency curve approaches the normal curve so closely. In 

 Fig. 8 is also drawn a strictly normal curve (log«//16 = — a;/3113). A 

 better fit could no doubt be obtained by suitable adjustment, but it 



