December, '12] HOLLOWAY: POISON EXPONENT 453 



order to obtain indicators which were even approximately correct it 

 would have been necessary to conduct all experi-ments in a place where 

 temperature and humidity were unvarying or varj-ing always in the 

 sartie degree. The pupation of the unpoisoned larvae complicated 

 the matter further. In outlining the experiments the time of pupa- 

 tion of a larva was arbitrarily taken as the end of an experiment, but 

 this procedure is of course open to objection. 



The experiments were discontinued and a new outline'^ of work was 

 prepared. This outline took Paris Green as the standard poison, and 

 larvae subjected to it were taken as checks. Paris Green, besides 

 being universally known, kills the larvae very quicklj', so that the entire 

 life of a check larva is of use in determining the toxic value of a poison. 

 The original outline had called for experiments with many different 

 dosages of the different poisons, but because of other work it was nec- 

 essary to adopt a uniform dosage of two milligrams per leaf, which is 

 equivalent to a rather thorough field poisoning. 



The idea of the toxic value of a poison was also changed. The pres- 

 ent conception may best be explained by algebraic symbols. 



Let X = an unknown poison. 



Let Paris Green = 1, or standard in toxic value. 



Let us suppose that an insect subjected to x dies after 10 hours, 

 and that an insect subjected to Paris Green dies after 5 hours. 



» The outline as finally adopted is in part as follows : 



Paris Green Lavanburg will be taken as the standard, and the larvae subjected to this 

 poison will be known as checks. 



The Poison Exponent for any poison will be obtained by comparing the length of 

 hfe of the larvae subjected to that poison with the length of life of a larva subjected 

 to Paris Green Lavanburg at the same time. Duplicate experiments will be con- 

 ducted if sufficient larvae are available. 



The poisons will in every case be evenly applied to the leaves in the form of a dust. 

 Two milligrams per leaf will be appUed. 



Experiments will be conducted with tin boxes as cages, a two-ounce size for the 

 larger larvae and a one-ounce size for the smaller larvae. 



One larva to a box. 



Larvae will be given young cowpea leaves of as nearly a uniform size as possible. 



After forty-eight hours' exposure to poison in each case, unpoisoned food will be 

 provided until death. 



Notes to be made at bi-hourly intervals of feeding in each cage. 



Records to be kept on cards, devoting one card to a lar\-a. 



For duplicate experiments, cards will be numbered with letters following the regu- 

 lar numbers as la, 2a, etc., for the first set of duplicates, lb, 2b, etc., for the second 

 set, and so on. 



The experiments will fall into two series: 

 Series I. Tests with all instars. 



Series II. Tests with the most resistant instar as determined by the tests in 

 Series I. 



