18 INSECTICIDES AND FUNGICIDES. 



per cent, and 1 has between 3 and 4 per cent, making a total of 19, or 

 43 per cent, which have more sodium sulphate than should be allow- 

 able. Of course, it is understood that this compound does not injure 

 the green in any way for the purposes intended, but adds weight, and 

 causes a high price for a cheaper article. 



The figures for total arsenious oxid show that no green contains 

 less than 56.20 per cent. There are 7 greens containing between 56 

 and 57 per cent, 24 containing between 57 and 58 per cent, 11 con- 

 taining between 58 and 59 per cent, and the remaining 3 containing 

 between 59 and 62 per cent. These figures would seem to indicate 

 that there is no reason for a sample of Paris Green containing less 

 than 56 per cent of total arsenious oxid, and that the States having 

 laws regarding this subject should change them so as to require the 

 presence of 56 instead of 50 per cent of arsenious oxid. 



The figures for copper oxid vary from a minimum of 27.58 to a 

 maximum of 31.16 per cent. More than half of the samples contain 

 between 30 and 31 per cent. 



If the Eastern and California standard of 4 per cent be adopted for 

 free arsenious oxid, and the 10-day water-extraction method be used, 

 there are only 13 samples, or 29 per cent, of the above Paris greens 

 that would pass, but if the 6 per cent limit, which we consider the 

 better, be used, 38 samples, or 84 per cent, of the greens would be 

 accepted. Of the others, 3 samples contain between 6 and 7 per cent, 

 1 sample between 3 and 9 per cent, 2 samples between 9 and 10 per 

 cent, and 1 sample 10.53 per cent. 



A column of figures is given, representing the amount of copper 

 oxid going into solution along with arsenious oxid when the green is 

 treated with water for 10 days. At present we do not know the rela- 

 tion between the amount of copper oxid thus extracted and the amount 

 of Paris Green broken up, so these figures can not be used to correct 

 the soluble arsenious oxid for the amount corresponding to Paris Green 

 broken up. These figures do, however, to some extent, seem to indi- 

 cate which samples give high figures for arsenious oxid because of 

 the breaking up of the Paris Green, and which give high figures because 

 of the presence of free arsenious oxid in the original sample. Hence, 

 the copper oxid determination seems to help us judge of the stability 

 of the green in question. For example, we would say that samples 

 19526, 19533, 20359, and 20365 give high figures for free arsenious 

 oxid because the green is unstable or is very fine, while samples 19712, 

 19718, 19726, and 20368 give high figures because the green contains 

 more free arsenious oxid originally than the average sample. This 

 course of reasoning is to some extent justified by the figures obtained 

 by the sodium-acetate-extraction method, which is supposed to indicate 

 more or less closely the amount of free arsenious oxid originally 

 present in the green. For samples 19526, 19533, 20359, and 20365 the 



