352 



later considered by him to be a distinct species 

 still imdescribed.) 



Two Cadicera hiclausa Loew (my Nos. 3,683 and 3,684, 

 sent to Guy Marshall). 



" The ñfteen butterflies that were swallowed, wings and all, 

 were : 



Seven Mycalesis campiña Auriv. 



Two Neptis agatha Stoll. 



Three Py rameis cardui L. 



One Precis orithya L. v. madagascariensis Guen. ( — 



boopis Trim.). 

 One Papilio demodocus Esp. (hindwings only : head 



and front of thorax also not swallowed). 

 One Rhopalocampta libaon H. H. Druce. 

 " The eleven that were eaten without wings were : 

 One Precis cehrene Trim. 

 Three Precis octavia Cram. var. geogr. natalensis Stand. 



dry season f. sesamtis Trim. 

 One Precis tugela Trim. 

 Two Precis artaxia Hew. 

 Two Py rameis cardui L. 

 One Pscudacrœa lucretia Cram. 

 One Char axes pollux Cram. 



" You have the entire pellet (I lost at the most only a few 

 scales in examining a portion of it), yet I doubt whether you 

 could easily ñnd in it evidence of the bird having shortly before 

 eaten all these Lepidoptera and Diptera ; yet the pellet repre- 

 sents the whole evidence (unless the intestines were also ex- 

 amined) that a man who shot the bird shortly before it was 

 ejected would have had to go on. 



" Small grasshoppers had been eaten just before the ex- 

 periment, and are also represented in the pellet. I also send 

 another smaller pellet of the same date, containing grasshopper 

 débris. I presumably kept it to compare with the one con- 

 taining butterflies. I have no note to show whether it came 

 before or after the latter. 



" I have practically never troubled to keep the pellets after 

 a butterfly experiment, but might easily do so, should you think 



