104 



G. HARKER. 



In distilling the sample oil, the residue is always measured 

 in order to provide a check against possible losses in dis- 

 tillation. It sometimes occurs that the distillation loss 

 exceeds 2 or 3 c.c., in this case the distillation must be 

 repeated. 



A comparison of the distillation figures for the sample 

 and blank leads to the conclusion that 19 c.c. of benzene 

 and 20 c.c. of toluene were present in the 68 c.c. washed oil, 

 or 15*2 per cent, of benzene and 16*0 per cent, of toluene 

 in the original No. I oil. The paraffins etc., contained in 

 the benzene and toluene were estimated by sulphonating 

 5 c.c. of the mixed benzene and toluene distillate; 0*2 c.c. 

 of oil was left equivalent to 4 per cent, of paraffins. This 

 proportion of impurities of uncertain boiling point is too 

 small to effect to any extent the comparison of the distill- 

 ation figures of the sample and blank. 



No. II Sample. 





Blank consisting of 





53 c.c. of washed oil. 



3 c.c. benzene, 12 c.c. 

 toluene, 8 c.c. xylene, 

 and 30 c.c. residue. 



Beniarks. 



71° first drop 



87° first drop 





80 0-5 c.c. 





Proportions for 



85 1-0 

 90 2-0 





blank test badly 



95 3-0 



20 c.c. 



estimated owing to 



100 





large ratioof higher 



105 50 

 110 6-0 



3-0 

 3-5 



boiling constituents 



117 11-0 



55 



to the lower boil- 



130 180 



11-0 



ing. 



140 23-0 



16-0 





Residue 300 







It is clear that the benzene in the sample must have 

 been between 4 and 4*5 c.c, and since 15 c.c. (equal to 

 volume of the 3 c.c. benzene and 12 c.c. toluene added) were 

 distilled over from the blank between 130° and 140°, it is 



