DETECTION OF ADULTERATION. 31 



and its adulterants give a precipitate on treating with a quantity of 

 saturated bromin water, and, as a consequence, but little reliance 

 can be placed on this test. One of the most reliable indications of 

 adulteration is the color of the dried sample. If lentiscus is present 

 it will darken greatly on heating, becoming a dirty light brown with 

 a tinge of red, while pure sumac only turns a slightly darker yellow. 

 The experienced analyst, having a pure sumac for comparison, can 

 pick out in nearly all cases samples adulterated with lentiscus. That 

 this test agrees well with the microscopical examination is shown by 

 the last column of Table II, where the purity of the samples as indi- 

 cated by the color after drying is given. Of a total of 91 samples 

 examined, 82 agreed with the microscopical test, 3 were doubtful, 

 and 5 were erroneous. Moreover the color of the extract has been 

 found a valuable indication, samples adulterated with lentiscus giv- 

 ing a dark reddish extract, easily distinguished from pure sumac. 

 As a rule, therefore, the experienced analyst can distinguish by 

 means of the color of the extract and the dried material those sam- 

 ples which are adulterated with lentiscus, but if there is any uncer- 

 tainty a microscopical examination must be made. 



Neither the percentage of ash nor of sand is an indication of adul- 

 teration with lentiscus, as this leaf does not differ materially from 

 sumac in these particulars. The samples of leaf and ground sumac 

 contained on an average 1.41 per cent of sand, the highest amount 

 found being 3.05 per cent. Assuming that there was no sifting out 

 of sand in transit, there was no evidence of willful addition of sand 

 to these samples, although several indicated that they were but 

 imperfectly winnowed or ventilated. There was less than 1 per cent 

 of sand in the unground leaf, while 106 samples of ground leaf aver- 

 aged 1.62 per cent and 15 contained more than 2 per cent. Therefore 

 2 per cent of sand may very properly be considered the maximum 

 sand content of a well- ventilated ground sumac. A larger content 

 of sand indicates that the samples have been carelessly prepared. 



117 



