183] 



THE GOLDFISH AS A TEST ANIMAL— POWERS 



63 



kleed has been very severely criticized. Many have questioned, as some have 

 put it, " the smooth and elegant results. " These smooth and elegant results 

 obtained by these workers are due at least in part to the fact that they were 

 working with solutions which fall within the portion- of the velocity of fataUty 

 curve that approaches a straight hne. (Compare to the portion A to B, 

 curve CABG, Fig. 1.) While the concentrations with which their critics 

 worked possibly fell outside this range of concentrations, it is probable that if 

 their data be reviewed from this point of view it will become more inteUigible. 

 From data obtained in this investigation it is clear that all experimentation 

 for pharmacodynamic assay work should fall within the portion A to B of 



TABLE XXXII 



The Variation of the Survtcval Time of Goldfish 



*Fish was not dead when taken out of solution. 



the velocity of fatahty curve CABG, Fig. 1. The question then arises by 

 what method can one determine with the least number of preUminary experi- 

 ments the portion of the curve at which one is working. This can be done 

 either by knowing the extremes of the survival time of the goldfish when the 

 concentrations are within the designated range (A to B) or by running only 

 three preUminary experiments at different concentrations. If the velocity 

 of fataUty curve formed by plotting the data from these experiments is convex 

 with respect to the X-axis, the substance should be tested by using solutions 

 falling near the weakest or between the two weaker solutions. If the curve 

 is concave, solutions nearer the strongest or between the two stronger solutions 

 should be used. If the curve is nearly straight and approaches a parallel 

 position to the X-axis the solutions are either too weak or too strong. (These 



