360 O. C. Farrington — Studies of Brazilian Favas. 



In the closed tube it decrepitated and gave off water. In 

 luster and specific gravity it corresponded closely to the other 

 favas and its color was light ochreous yellow. After fusion 

 with JNa 2 C0 3 this fava dissolved easily in H CI to a yellow solu- 

 tion. Analysis of this by Mr. H. W. Nichols gave the follow- 

 ing result : 



A1 2 3 61-67 



Fe 2 O s 4-37 



BaO 2-66 



Si0 2 10-40 



S0 3 0-96 



P 2 5 0-76 



H 2 17-66 



Moisture 1-60 



100-08 



The mineral is thus seen to be essentially hydrous oxide of 

 aluminium. Regarding alumina and water as the essential 

 ingredients and calculating to 100, the analysis gives : 



A1 2 3 77-74 



HO _ 22-26 



100-00 



This corresponds closely to the formula 2A1 2 3 .3H 2 0, the per- 

 centages of which are : 



A1 2 3 79-1 



HO 20-9 



100-00 



Here again a new mineral is perhaps indicated, but owing to 

 the impurities of the material the writer is not inclined to 

 urge this conclusion. The chief result which seems to the writer 

 to be indicated by his investigation of this group of favas, 

 is that even among favas closely resembling each other in 

 color, texture, luster and size there may exist substances of 

 very different chemical composition. It is probable that a 

 similarity in specific gravity was the chief agent in bringing 

 into close association substances of such varying composition. 



Field Museum of Natural History, 

 Chicago, 111. 



