102 Prof. H. E. Armstrong on the Determination of 



(20) Thus far I have only endeavoured to give an abstract 

 of Thomsen's arguments and conclusions, quoting as far as 

 possible his own words. Many of his deductions are in the 

 highest degree remarkable. Recalling the more striking, he 

 not only finds that the same amount of heat is developed in 

 the combination of carbon atoms as they occur in ethylene 

 and as in the paraffins — in other words, that even in the ole- 

 fines the carbon atoms are united only by single bonds — that, 

 in fact, there are no such things as " double bonds "; but he 

 also arrives at the startling conclusion that in the formation 

 of acetylene the carbon atoms unite without any evolution of 

 heat : so that we are forced to assume that in an acetylene 

 not only is there no treble or double bond, but not even a 

 single bond ! Then ethylene oxide is pronounced to be methy- 

 lene oxide ; the aldehydes are /w/iiro^-compounds ; and pro- 

 bably methylal and methylic orthoformate are also alcoholic 

 bodies. Lastly, the amines are to be regarded as derivatives of 

 pentad nitrogen ; and pyridine is not analogous in constitution 

 to benzene. 



(21) Now, admitting even that a more careful consideration 

 of the chemical evidence might result in our acknowledging 

 the correctness of Thomsen's conclusions in some few cases, 

 it is impossible to do this in the majority of instances : we 

 cannot admit that the carbon atoms in ethylene oxide are dis- 

 united, and that this compound is in reality a methylene oxide; 

 and every chemist must regard Thorn sen '$ formulae for the 

 amines as altogether lacking probability. Moreover, the 

 method by which the constitution of bodies like ethylene oxide 

 and the amines is ordinarily arrived at, is the method by 

 which the constitution of compounds generally is determined. 

 Hence, if we accept Thomsen's conclusions in their entirety, 

 results arrived at by the strict application of the same method 

 throughout are to be accepted in some instances but rejected 

 in others. To admit this would be to acknowledge that our 

 entire system of constitutional formulae is based upon a false 

 conception, to which there is no possible key. That current 

 views of structure require modification in some not unessential 

 particulars, I have long been of opinion ; but that they will 

 have to be modified to the extent which Thomsen's arguments 

 indicate appears to me altogether improbable. 



(22) It remains, therefore, to seek for some explanation of 

 his anomalous results, and one of the first questions to be 

 answered is: Are we justified in regarding the value 135340 

 units as the true heat of combustion of gaseous atomic carbon ? 

 Thomsen's determination of this value involves the assump- 

 tion : that when a molecule of oxygen combines with two 

 molecules of gaseous carbon monoxide, the same amount 



