February 28, 1908] 



SCIENCE 



323 



(hg = ^ an atomic weight of bivalent mer- 

 cury.) 



hg — — N = C^ ^hg — O — N 



o 



^H 





0- 



J! 



:C+ I 



OH 



The yellow salt formed at the same time 

 may be a basic mercury salt of formhy- 

 droximie acid. 



hg — O — N = C 



\o-Hg-OH 



This explanation was offered (Ann., 20, 

 34) in view of the fact that the yellow salt, 

 upon treatment with dilute hydrochloric 

 acid, yielded some fulminate. "Wohler 

 {Ber., 38, 1351) failed to obtain fulminate 

 from this salt. Recently, a repetition of 

 these experiments has yielded fulminate 

 repeatedly. 



If the reactions represented by the equa- 

 tions given above are assumed to be slightly 

 reversible, it is possible to account for many 

 of the singular changes of the fulminates. 

 Thus, the change of fulminate into form- 

 hydroximic acid would account for the 

 decomposition of fulminates into hydroxyl- 

 amine and formic acid. By the Beckmann 

 rearrangement, formhydroximic acid might 

 be expected to pass into isocyanic acid, or 

 into ammonia and carbon dioxide. 



H— o-]sr=c< 



/OH 

 .C< 3&-^ H— N 

 ^OH 



H 

 =C=0-|- I s-^NH3=C02. 

 OH 



This would serve to clear up the behavior 

 of fulminates towards acetyl chloride, am- 

 monia, aniline, thioeyanic acid and certain 

 other reagents. 



In 1856 Kekule represented fnlminie 

 acid as nitro acetonitrile, CHoCNOa) (CN). 

 The chief experimental evidence for this 

 formula lay in the fact that mercuric ful- 



minate, under the influence of chlorine 

 water, gave trichlornitromethane (chlor- 

 pikrine), and a small amount of chlor- 

 cyanogen. The formation of trichlornitro- 

 methane has never been explained; but its 

 formation becomes self-evident in view of 

 the above hypothesis, since mercuric isoni- 

 tromethane in the presence of mercuric 

 fulminate would be changed ultimately into 

 trichlornitromethane, free fulminic acid, 

 and mercuric chloride by the continued 

 action of chlorine water. The presence of 

 a small amount of chlor-cyanogen may be 

 accounted for by the presence of prussic 

 acid, a substance very often formed by the 

 breaking down of free fulminic acid. 



The Asymmetric Methyl Dialkylisoureas : 

 Ealph H. McKee, Lake Forest Univer- 

 sity. 



The isoureas of this series, AlkoN — 

 COCHj^NH, were prepared by the ac- 

 tion of methyl alcohol on the dialkylcyan- 

 amide in the presence of sodium methylate. 

 They are strongly basic oils whose water 

 solutions act similarly to ammonia on sil- 

 ver, mercuric and mercurous salts, but are 

 unlike ammonia in that they do not affect 

 copper or cobalt hydroxides. Decomposi- 

 tion of the dry hydrochlorides, or their 

 solutions in water, gives methyl chloride 

 and the urea (Alk^N — CO — NH^) quan- 

 titatively. These isoureas readily condense 

 with phenyl isocyanate, benzoyl chloride, 

 etc., but are not affected by benzaldehyde, 

 ethyl oxalate or ethyl malonate. 



Melting Points of Binary Mixtures of 

 Ortho-Meta and Paranitr aniline — A New 

 Method for the Determination of these 

 Compounds: J. Bishop Tingle and H. 



F. ROLKEK. 



Binary mixtures of the three nitranilines 

 have been made in steps of 2 per cent., and 

 the melting points of these mixtures have 

 been plotted against the composition. The 

 curves given by the ortho-meta and meta- 



