275 



of food. The native women make and sell great quantities of it 

 in the Asuncion market. In the town of Luque, on the Asuncion 

 and Villa Rica Railroad, they are noted for the fine bread which 

 is made of this flour, and the passengers eagerly purchase it from 

 the women who offer it for sale. I have also seen pastry and 

 sponge cake made of the flour as light and palatable as anything 

 prepared from wheat flour, and I do not know why it may not 

 serve all the purposes for which the latter is used. 



If the ground powder is heated upon iron plates and partially 

 cooked, it clusters into hard and irregular lumps, and forms the 

 well-known tapioca of commerce, or prepared somewhat differ- 

 ently it becomes the article known in England as '* Brazilian ar- 



row-root" 



The Mandioca brava does not differ much from the NL didce 

 in external botanical characters. Both are stout herbs, growing 

 from five to eight feet high, branching and very foliaceous. Both 

 have their flowers in short axillary racemes, the flowers small, 

 purplish-white and nodding, producing a capsular, 5-valved fruit, 

 and an acrid, milky juice. The leaves of both species are alter- 

 nate, on long petioles, palmate, with narrow, deeply cut lobes. 

 The natives, however, readily distinguish the two species. M. 

 duke, they say when questioned, has red stems, petioles and 

 leaves, while those of M. brava are white. On examination, 

 however, it will be found that this distinction will not always 

 hold good, as M, ^///r*? frequently has stems and petioles almost 

 or quite white, while those of J/, brava are not unfrequently red- 

 dish. A much better distinction lies in the shape of the stems, 

 which in the former case are nearly^ or quite terete, while in M. 

 brava ^i\\Qy are more or less angled; and also in the angle at 

 which the petioles of the two species spring from the stems, ris- 

 ing in M. dtilce at a right angle, or even sloping downwards, 

 and in M. brava at an angle of sixty or forty-five degrees. After 

 all, the only satisfactory distinction lies in the juices of the two 

 plants. That of M. duke, as already stated, is sweet and innoc- 

 uous. That of M, brava, on the contrary, is poisonous. The 

 juice of this species has been known to kill cattle if the roots are 

 eaten in the raw state. If, however, the juice is thoroughly 

 squeezed out, and the grated pulp dried, it may be used in mak- 



