10 COMMERCIAL SICILIAN SUMAC. 



To determine the sand, treat the ash with about 10 per cent hydro- 

 chloric acid and warm gently for several hours, filter, wash thor- 

 oughly, ignite, and weigh. The weight so obtained is considered as 

 sand. 



Make the color determination of the extract with the Lovibond tint- 

 ometer, the readings being made in the 1-inch cell on the soluble 

 solids filtrate, and calculate the results to a basis of 0.5 per cent of 

 tannin in the solution. Give the results in terms of red and yellow, 

 .the black being subtracted for the reel and yellow readings. 



It should be borne in mind that the color determination s6 obtained 

 does not necessarily represent the color of the extract obtained in 

 tannery practice. Indeed, it is almost certain that the color of the 

 extracts as thus prepared for analysis is considerably deeper than 

 that of the extract obtained from the same materials by ordinary tan- 

 nery methods. These results, then, only show the relative colors 

 produced by different samples under like conditions of extraction 

 and should be compared only among themselves. 



DISCUSSION OF RESULTS OF ANALYSIS. 



In tabulating the data on these samples, the name and address of 

 the consignor, when these appeared on the containers, the name and 

 address of the consignee, the approximate date of entry, and the 

 place of sampling are given in connection with the chemical analysis 

 and the microscopic examination. Care was taken to secure sam- 

 ples only from invoices imported in good condition. 



As is shown in Table II (p. 12), the average percentage of tannin 

 in all the samples of sumac is 28.8 per cent, which is higher than the 

 results generally given in the literature for Sicilian sumac. As has 

 been said, it is not believed that this is due to any improvement in 

 the character of the leaf now grown, but rather to improved methods 

 of extraction and also to changes in methods of analysis. 



Approximately 41 per cent of the invoices from which samples 

 were taken were mixed with lentiscus, this being practically the only 

 adulterant employed, except sumac stems, which were present in 

 excessive quantities in a number of samples. The adulterated sam- 

 ples contained from 19.6 per cent to 33.3 per cent and averaged 26.6 

 per cent of tannin, or 2.2 per cent less than the average of all the 

 sumac samples. A number of the samples contained an excess of 

 sumac stems, and the average tannin content of these was 29.9 per 

 cent, which indicates that the stems are not added in such large 

 quantities as is the lentiscus. The samples of pure sumac contained 

 from 27.4 to 35.1 per cent and averaged 31.9 per cent of tannin. 



Adulteration of the average pure sumac with 30 per cent of the 

 average lentiscus would yield an article having the same percentage 



117 



