﻿Critical Temperatures of Homologous Compounds. 607 



It is of interest to see whether analogous formula? can be 

 applied to those series in which the influence of association 

 is manifest — the primary alkyl alcohols, for example. If 

 the graph between logo and log n be plotted for this series, 

 it will be seen that whilst the relation between log p and 

 log n is still approximately linear, the points are scattered 

 rather irregularly about the line passing through their mean 

 position. Calculating for this line, we find that 



6> A = 0P-f 



331-1 



1 -/7456 



(xvii.) 



giving the boiling-point of the alcohol (#a) in terms of that 

 of the corresponding paraffin (Op). The results are shown 

 in Table V. 



Table Y: 



Alcohol. 



G? obs. 



0a ubs. 



6a calc. 



Difference. 





o 



o 



o 



o 



C 3 H 7 OH .. 



228-0 



370-2 



374-0 



4-3-8 



C 4 H 9 OH ... 



274-0 



3899 



391-8 



H-9 



C.H n OH ... 



309-3 



4110 



4090 



-20 



C 6 H 13 OH ... 



342-0 



431-0 



429-0 



-20 



<XH 13 OH ... 



371-4 



4490 



449-0 



+o-o 



8 H 1T OH ... 



398-6 



464-0 



468-8 



4-4-8 



C n H 10 OH ... 



422-5 



4865 



486-8 



4-03 



C 10 H ol OH... 



446-0 



504-0 



505-5 



4-1-5 



Average difference regardless of sign =2 o, 04. 



Whilst, therefore, the equation for the bromides is in close 

 agreement with the observed facts, that for the alcohols shows 

 much greater divergences. A similar divergence has been 



O CD O 



noted by Young, who, using equation (iii.), has calculated 

 values for A for each of the series considered above. In the 

 case of the bromides the average difference regardless of 

 sign between the observed and calculated values of A is, 

 over the range taken, 1 0, 25. In the case of the alcohols the 

 average difference between the observed and calculated 

 values of A is 4 0, 28. Thus, while the average values of the 

 errors in A are considerably larger than those in the form uke 

 (xviii.) and (^xvii.),, the ratios of the errors are very approxi- 

 mately the same, for in each case the error for the alcohols 

 is about three times the corresponding error for the bromides. 



* Observed data from Young, /. r. 



