December, '15] 



WOODWORTH: THEORY OF TOXICITY 



511. 



in any similar investigation and gives results of a very dependable 

 character. The methods of study have already been described else- 

 where/ and it will only be necessary to explain that this study presents 

 the result of 200 sets of experiments, each set consisting of simulta- 

 neous determinations for each of the 100 points on the diagram, made 

 with special care to insure that all the data in each set shall be com- 

 parable. 



Each curve represents the percentages of the lots which failed to 

 hatch after treatment in a certain density of gas through exposure for 

 varying intervals of time. These intervals range from five minutes 

 to five and one quarter hours. The intervals indicated by the vertical 

 lines are in geometric progression. The horizontal lines represent the 

 number of lots giving no hatch. 



The strength of dose doubles towards the upper end of the table 

 for each line, the topmost line being 512 times as strong as the weak- 

 est, which is dotted on account of its somewhat erratic position. 



Seven theoretical curves are drawn which correspond very closely 

 with the experimental data and show that very complex relationships 

 exist. It is likely that the same facts will be found true for other 

 insecticides and that we can establish, at least as a working hypothesis, 

 the series of toxic phases shown on the diagram, with the definite 

 relationships there indicated. These phases are as follows: 



The Preliminary Curve in which the toxicity is directly propor- 

 tional to an increasing geometric series of time intervals, and for the 

 stronger doses, is also directly proportional to an increasing geometric 

 series of concentration intervals. 



The Deviation Line is the center of a very evident zone where 

 the toxic curves diverge. It seems to be inversely proportional to 

 the toxicity during the preliminary period; that is, the less the killing 

 the longer the interval, the increase of time corresponding with a de- 

 crease of dose being in geometric ratio and, therefore, making a 

 straight fine where the time intervals are plotted as in this chart. 



The Acute Curve represents a different physiological action of the 

 cyanid, possibly directly upon the nerve centers. It is a more abrupt 

 and much longer continued action in the lower concentrations of the 

 poison, resulting in a far greater difference in the time required to 

 complete the action than in percent that finally dies. The propor- 

 tional rate of kilhng during this acute phase is identical for all con- 

 centrations, the greater kilhng for low concentrations depending solely 

 upon the longer time of action. 



1 Science, Vol. xli, pp. 267-269, and Report California Agricultural Experiment 

 Station for 1914, pp. 114-116. 



