12 BULLETIN 1160, U. S. DEPARTMENT OF AGRICULTURE. 
compounds when used as fumigants, but state that at a certain point 
in the series the toxicity becomes uncertain because of the small 
amount of the compound which enters the gaseous phase owing to 
the low vapor pressure. This is certainly not true of the aliphatic 
substituted ammonium salts mentioned above when used as contact 
insecticides, since they are practically nonvolatile and at the same 
time quite toxic. 
Tattersfield and Roberts indicate that the effect of substitution of 
various atoms and groups into a compound depends upon the nature 
of the parent compound and the group or atom introduced. Some 
of the writers' results are in harmony with theirs. Thus they found 
that each CH 3 introduced into benzene approximately doubled the 
toxicity of the substituted compound. Table 2 shows that benzene 
was toxic at 3.2 moles per liter, toluene (C 6 H 5 CH 3 ) at 1.7, and xylene 
(C 6 H 4 (CH 3 ) 2 ) at 0.94, or nearly double the toxicity for each CH 3 group 
introduced. Moore 05) found benzene to be slightly more toxic than 
toluene in the gaseous phase against house flies (Musca domestica L.). 
Xylene, however, was more toxic than either. The chlorine atom, on 
the other hand, was found by Tattersfield and Roberts to increase the 
toxicity of the substituted benzene compounds from three to four 
times for each substitution. A similar result was obtained by us 
with chlorobenzene, the molar toxicity of which was 0.80, or four 
times that of benzene. Commercial trichlorobenzene was more toxic 
than chlorobenzene, but not in the same ratio. It was possibly 
a mixture of several isomers. Previous work by Moore (5) shows that 
the introduction of halogen into the benzene ring increases toxicity, 
iodine being the most effective, chlorine the least, and bromine having 
an intermediate value. The writers' experiments show a greater 
toxicity when OH is introduced into toluene to form cresol than when 
introduced into benzene to form phenol, a result which is the reverse 
of that recorded by Tattersfield and Roberts for wireworms. 
Introduction of NH 2 , OH, or CI in the side chain (CH 3 group) of 
toluene gave interesting results. Benzylamine is the most toxic, and 
benzyl alcohol the least, when the molar toxicities are compared. All 
are considerably more toxic than the parent compound, toluene. The 
experiments of Tattersfield and Roberts indicate that benzyl chloride 
in the vapor phase is highly toxic to wireworms. 
Moore (5) has pointed out the increased toxicity resulting from the 
introduction of the CHO group into benzene, and results obtained i 
in the present study with benzaldehyde as a contact insecticide against 
Aphis rumicis indicate a similar relationship. 
Tattersfield and Roberts observed considerable toxicity among the 
aliphatic amino compounds which they tested. Methylamine, dime- 
thylamine, and ethylamine were about as effective as fumigants for 
wireworms as hydrocyanic acid and ammonia; trimethylamine was 
slightly less poisonous. They were far more toxic than pyridine, 
and somewhat more toxic than aniline. Their toxicities did not in 1 
any sense correspond with their extremely high volatilities. It was 
suggested that tneir solubilities, the readiness with which they are 
absorbed, and their ability to ionize may account for the relatively 
high toxicities of these compounds. The writers have observed high 
toxicity among certain aliphatic amines and related compounds. If 
the results with triethylamine may be taken as representative, the 
salts of these amines are more effective than the corresponding bases. 
