Dr. G. D. H. Carpenter on Pseiidacraen eunjivs hohleyi. 85 



Lastly, after the signing of the armistice I was enabled 

 to return to work on the islands, and went back to Kome 

 in Dec, 1918, where I obtained more specimens from the 

 same localities as in 1914, and also from the eastern end 

 of the large island of Kome. 



The specimens obtained from these localities have 

 afforded most interesting and important data, exactly 

 complementary to those in the earlier paper. The figures 

 given in that paper may be here reproduced, together with 

 the new data now discussed. • 



Mainland, 1909 (C. A. Wiggins, Entebbe) 



Bugalla Island, 1912-13 



Kome and neighbouring islands, 1914 



Mainland (Kakindu), 1915 



Kome and neighbouring islands, 1918-19 



Sundry other islands, 1918-19 



Study of the present data shows that : — ■ 



I. In 1914 on the islands Planema models were found 

 to be very numerous and the Pseudacraea mimics 

 less abundant and true to type, as in the Wiggins 

 collection from the mainland. 

 11. On individual islands the predominant form of 

 Pseudacraea was that which mimics the pre- 

 dominant Planema of that locality. 



III. The collection from the Tero forest on the mainland 



(Kakindu) confirms the last result. In this locality 

 the only abundant form of Planema was copied 

 by the appropriate form of Psevdacraea, while 

 only a single specimen of another form was 

 taken. 



IV. On the same islands in 1918-19, Pseudacraea was 



again more numerous, and more variable. 

 V. A collection from sundry other islands in 1918-19 

 contains ten times as many Pseudacraea as 

 Planema models, nearly half of which are tran- 

 sitional. 

 The total numbers for 1914 are tabulated below for 

 comparison with the Bugalla captures : the former are 

 indicated by ordinary figures, the latter by Roman figures. 



* The total was Avrongly added up to 127 in the description of 

 the table in tlie 1914 paper. 



