Affinity and Dissected [Structural) Formula, 



427- 



rections, it may fairly be concluded that the directly calculated 

 values are erroneous to the extent thereby indicated. 



The wonder is, not that a few of the affinity- values calculated 

 as above from data often defective or approximative only should 

 be clearly somewhat erroneous, but that so large a number of 

 instances should be found indicating* that the reduced affinity- 

 values thus obtained are not far removed from the actual true 

 values, and that it should be possible from experimental deter- 

 minations so incomplete to deduce general rules of so wide an 

 application and subject to so few apparent exceptions. 



40. If Favre and Silbermann's determinations of the heats of 

 combustion of various terpenes could be thought to be absolutely 

 exact, these substances would form a well-marked exception to 

 the thermal rule above stated. 





Boiling-point. 



Value of H r 



Affinity- value. 



Terebene 



156 

 161 

 173 



10-662 

 10-852 

 10-958 



A. 



Less than A. 

 Still smaller. 



Oil of turpentine ... 

 Oil of lemons' 







So that rise in boiling-point would be accompanied by diminution 

 of affinity-value. But it has been shown by the writer, and also by 

 Orlowski, that distilled oil of turpentine naturally contains cy- 

 mene; the same remark probably also applies to oil of lemons; 

 whilst Riban has shown that the process ordinarily used for pre- 

 paring terebene from turpentine-oil (by means of sulphuric acid) 

 gives a product containing much cymene. Hence in all proba- 

 bility the bodies examined by Favre and Silbermann were not 

 pure homogeneous liquids ; and therefore no conclusion can be 

 drawn from the difference in their heats of combustion. 



41. In addition to the foregoing conclusions, some others may 

 be drawn from the affinity-values calculated above. Thus : 



When two substances have formulas differing only by C sym- 

 bols, the one with the higher number of C symbols has a less 

 affinity-value than the other. 



Affinity-values. 



Marsh-gas ... 21-7 



Alcohol 63-7 



Acetone 59-9 



Alcohol 637 



Ethylene... — 55 



Acetone ... -j-59*9 



Phenol +29'5 



Phenol +295 



Number of additional 

 C symbols. 



Difference of 

 affinity-values 



—•27-2 



- 3-8 

 -30-4 

 -34-2 



When the formulas differ only by H symbols, it would seem 

 that the one with more H symbols has a higher affinity -value than 

 the other. 



