69 



was given to this office, and a stronger test AA^as recommended, three 

 times the usual strength — that is,, 3 ounces of cyanide of potash to 

 100 cubic feet of space — and a longer exposure. This was applied 

 September 28, and renewal was made the following day, twenty- 

 three hours later, arrangements having been made so that jars con- 

 taining fresh acid could be introduced and charged with cyanide 

 through a window. The second exposure lasted nineteen hours, or 

 practically forty-two hours in all. 



Prior to experiment it was ascertained that the insects had been 

 breeding in the loAver coA'ering of the chairs. Dozens of larvae were 

 found betAveen the covering and the Avebbing Avhich was nailed across 

 the latter. 



After airing the infested room hundreds of larvae were observed 

 that had fallen from the lower covering of chairs (which had pre- 

 viously been ripped off to facilitate the effect of the gas) to the floor, 

 and all adults observed Avere dead, as Avere also house flies. Some of 

 the larvae and beetles Avhich had been subjected to treatment were 

 placed in a vial to determine the effect of the experiment. They all 

 died, but considerably later it Avas found necessary to dispose of the 

 chairs, as they Avere still infested. 



OAving to the failure of our first experiment with hydrocyanic-acid 

 gas against the cigarette beetle, as also of another experiment on the 

 confused flour beetle {Triholium confusum Duv.), it was suspected 

 that something might be wrong Avith the ingredients, and accordingly 

 a sample of cyanide of potash Avas selected at random and submitted 

 to Mr. J. K. HayAvood, of the Bureau of Chemistry, for analysis. He 

 rejDorted that it contained 37.18 per cent c3^anogen, 9.57 per cent 

 chlorine, and the remainder a mixture of potassium and sodium, the 

 analysis showing that this sample was not the product paid for; in 

 other words, not 98 per cent potassium cyanide, but a mixture of 

 potassium cyanide, sodium cyanide, and sodium chloride. The three 

 compounds were present in such proportions that the mixture yielded 

 93 per cent of the amount of hydrocA^anic-acid gas it should yield if 

 the sample were pure potassium cyanide, and yet the amount of cyan- 

 ogen AA'hich might be prodacecl by this mixture could be 53 per cent, 

 Avhereas in pure potassium cyanide it is 40 per cent. 



In the second test against this species the potassium cyanide was 

 used three times as strong as in the first. In the meantime, a sample 

 was being analyzed by Mr. J. K. PlayAvood. His results Avere as 

 f olloAvs : 



Per cent. 



Potassium cyanide 51. 70 



Sodium cyanide 2. 07 



Sodium chloride 3.89 



Potassium carbonate 39.28 



Other impurities 2. 70 



Moisture 0. 30 



