— 28 — 



an desiccator, and weighed. Non-volatile impurities can be determined 

 in the smaller watch-glass. The influence of the time during which 

 the heating takes place is not of much importance, as is shown by 

 the following data taken from the detailed tables of the authors: — 

 After heating for 10 minutes, found 97,05 °/ camphor (mean of 4 tests) 

 h 20 „ „ 9 6 > 86 % » » » 2 » 



» 3° » » 97.04 % » » » 3 » 



» 45 » >> 9 6 > 8o °/o n 11 »i 2 » 



., 60 „ „ 9 6 > 68 °/o » ( one test )- 



The results therefore agree within +o,5°/ ; the correction required 

 is -J~"3i°%- This method is useful for rough but quick determinations. 



2. Specific gravity. First of all tables 1 ) were calculated, which 

 give the specific gravities of solutions of the purest camphor in pure 

 commercial benzene (of E. Merck), namely o to 30 g. camphor 

 increasing by one half gram, dissolved in benzene up to 100 cc. 

 Next, exactly 10 g. of the sample of camphor in question were dis- 

 solved in pure benzene in a 100 cc. graduated tube, and filled up 

 to 100 cc. From the specific gravity, determined pyknometrically, like 



20 



the figures given in the tables, at — , the camphor content was 



I5i5 

 calculated by means of the tables. 



3. Polarisation. The optical method indicated by Landolt 

 and modified by Forster, consists of the determination of the rotation 

 of a solution of camphor in benzene, in Laurent's apparatus. The 

 calculation is made according to the formula 



-j/. 



0,04367 • a 

 - 1 



in which c represents the number of grams camphor of 100 cc. solution, 

 a the rotation, 1 the length of the tube in decimetres. The normal 

 temperature, 20 , must be maintained exactly. The camphor was 

 dissolved directly in an (American?) commercial benzene, as in the 

 pyknometric determination, without previously separating the non- 

 volatile substances according to Forster; but the figures obtained 

 were 2 to 3 °/ higher than those expected theoretically, because the 

 benzene (as was found on examination) contained considerable quantities 

 toluene and thiophene. With pure Merck's benzene the correct values 

 were obtained, both with a Laurent's apparatus and with one of 

 Schmidt and H a e n s c h. The deviations fluctuated between 

 + 0,07 and -f-0,14%. 



The method of A most, basing on the volumetric expansion of a 

 petroleum ether solution, was found to be useless. 



*) Reproduced ibidem p. 386. 





