1S92.J Elliot, TnhcritaJicc of Acquired Characters. 8 1 



of "sufficient precision" to prove the theory he advocates than 

 those w^ho as yet decline to accept his views can bring forward 

 as practically substantiating their position. Neither side has 

 proofs that can demand the pronouncing of a final verdict, and 

 we can consider only which evidence produced is most likely to 

 show the true solution of the problem. 



Some of the evidence of variation produced, as we believe, by 

 environment and isolation may be derived from various genera of 

 birds containing numerous species with a wide dispersion, and of 

 these I would cite Ptilopus. a genus of Fruit Pigeons. These 

 birds are chiefly inhabitants of islands, and intercourse between 

 them is impossible on account of the intervening sea. As an 

 example of variations that are produced, I will select ;is a type 

 Ptilopus melanocephaliis^ a thoroughly characteristic and well- 

 marked form, known to and accepted by ornithologists as a 

 species, and which is an inhabitant of Java, Lombock, Sumbawa, 

 and Sulabessie. In the neighboring island of Flores is a form 

 Ft. melanatichcn., varying but slightly from the type as if just 

 commencing to diflerentiate. To the north in the great island of 

 Celebes is another variation called Pt. inela>wspihis diftering in 

 a narrower throat mark, and in having the crissum lemon-yellow 

 washed with orange, instead of clear yellow. Sula and Ceram 

 possess another form which has the occipital black band (present 

 in all the forms cited) smaller, and the crissum orange yellow ; 

 and lastly in the island of Sanghir, farthest away, Pt. xanthor- 

 rhous is found, which has both abdomen and crissum orange, this 

 color running upwards nearly to the breast, thus exhibiting the 

 widest divergence from the t3'pe. It will be seen from this dis- 

 tribution that, as might have been expected, the variations in the 

 foims living nearest to the type, and consequently with less 

 change of environment, are the slightest, while as the separation 

 becomes greater these dilTerences are more pronounced, luitil in 

 Sanghir appears a form which has departed so greatly from the 

 type as to merit specific rank. 



The probable cause of these diflerentiations in the members of 

 Ptilopus may be explained as follows. The islands of the East- 

 ern Archipelago doubtless are but the remains of what was once 

 a continent, and this was not broken up simultaneously or always 

 suddenly in its length and breadth, but sometimes gradually and 

 at various periods. Therefore we should not be surprised that 

 11 



