552 THE POPULAR SCIENCE MONTHLY 



being sufficient to produce death per kilogram, while, as we have seen, 

 7.75 grams of ethyl alcohol were required. 



But the methyl alcohol of the time of Dujardin-Beaumetz and 

 Audige, as we now know, was far from pure — hence the failure to gain 

 an accurate measure of its toxicity. To-day that methyl alcohol is pro- 

 duced in greater purity, we should be able to retest the question with 

 greater accuracy. To this we shall soon return. 



From the investigations of Dujardin-Beaumetz and Audige we have, 

 then, our first experimental evidence that while alcohols in large doses 

 are poisonous, not all alcohols are equally poisonous. To them is also 

 due the credit of showing that for the alcohols of fermentation the 

 toxicity is directly in proportion to the molecular weight and boiling 

 point ; in other words, that they are in accord with the law of Eabuteau 

 — the higher the molecular weight and boiling point, the greater the 

 toxicity. 



Difficulties Confronting Investigation of Toxicity 



Concerning the difficulties confronting investigation in this subject, 

 we have said nothing. Some of these we shall now consider. 



Observation has been made by various investigators that different 

 animals react differently to poisonous substances. From this observa- 

 tion arose the discussion as to " the choice of animal " best suited to a 

 study of alcoholic poisoning. Morgan, who had experimented upon the 

 dog, argued that it was the most acceptable animal for work of this 

 sort. Laborde, who, on the other hand, had studied the guinea-pig, 

 urged the use of this animal; while Colin (who had studied neither) 

 was of the impression that the horse or the cow would be more suscep- 

 tible than either. Daremburg, in considering the discussion, hopefully 

 suggested that probably another contradictor would recommend either 

 the giraffe or the elephant. This in fact might have been the plight had 

 not the work of Joffroy and Serveaux appeared. 



Joffroy and Serveaux have shown that while animals differ in sus- 

 ceptibility according to their kind, this difference is relatively constant 

 and usually but slight. The choice of animals thus becomes, in great 

 part, important in so far as one animal rather than another serves 

 better the purpose of this or that investigator. 



A question of more than passing importance in the measure of toxic- 

 ity, however, is the method of administering the substance to be tested. 

 It is a well-known fact that some substances which are extremely poi- 

 sonous when injected under the skin, for example, snake venom, are in 

 no sense poisonous when given by the stomach. On the other hand, 

 other substances which show slight poisonous effects if given subcu- 

 taneously, act with extreme rapidity if added directly to the blood 

 stream. These facts give to the ways and means of injection a high 

 importance. 



