( 440 ) 



dissirailis-form — wliile on the Andamau Islands only a dissi'inilis-fonn and on t lie 

 Philippines only a ciytia-iorm occurs; both the Andaman and the Philippine tnono- 

 morphic insects are not specifically distinct from the Indian dimorphic insect. The 

 nearest ally of Papilio clylia is the very variable Papilio paradoxus* If we assume 

 that the Andaman and Philippine insects are the descendants of the Indian dimorphic 

 clytia, or the latter the descendant of the former, or all three the descendants of a 

 common ancestor, it is evident that the characters by which formerly (before the 

 separation into three geographical races) the species was distinguished from prira- 

 doxiis are not the same as those by which all forms from the three localities togetlier 

 (which forms now constitute the species) are at present distinguished from paradoxus. 



We may add a few more illustrations from otlier well-known butterthes. The large 

 Papuan butterflies usually called Omithoptera are in the tnaie sex of an orange, 

 green, or blue colour. For the sake of argument let us assume that the green 

 priainiis, orange croeaus and lydius, and blue urvillianus were derived from a 

 common ancestor which had an orange inale. The nearest ally is 0. tithonus, which 

 has a green nude. In the diagnosis of the gi'een ancestral forms of tithonus tlie 

 diS'erence in colour between it and the orange ally would have been a specific 

 difference ; the recent tithonus cannot be differentiated from imarnus by the shade 

 of colour, as both insects are green. This one specific character has not remained a 

 specific character in tithonus, though perhaps the ancestral tithonus and the recent 

 one are identical in colour, and though the character itself, therefore, has proved to 

 be hereditary; on the other side, that one distinguishing character of the orange 

 ancestor of pi'iaraus, croesus, lydius, and urvillianus has been inherited only by 

 lydius and croesus, not hj pinamus and urvilliaiius, and therefore has proved to be 

 only restrictedly hereditary. 



One of the characters by which Piipiiio eiirypylus is distinguished, for instance, 

 from P. sarpiedon, is the position and extent of the red costal mark on the underside 

 of the hiudwing. This mark is in eurypylus a small red spot before the costal 

 nervure, while in sarpedon the spot is extended beyond the costal to the subcostal 

 nervure. Quite recently Mr. Walter Eothschild received a series of specimens of 

 P. eurypylus from the Kei Islands,t among which is one that has the red mark 

 extended to the subcostal vein. The extent of this spot, which until now was a 

 specific character of eurypylus, is no longer of specific value, and cannot serve to dis- 

 tinguish eurypylus from sarpedon or viacfarlanei : in respect to the latter two 

 species the form of the spot is hereditary, but no longer s[)ecific ; in respect to 

 eurypjylus it is not unrestrictedly hereditary, and also no longer of specific value. 

 Now, if it should happen that all specimens of eurypiylus acquired the extended red 

 mark and became at the same time in other characters still more different from 

 sarpiedon and macfarlanei, would then carypylus, surpedon, and macfarlanei of 

 our daj's not be specificall}' different, because their offspring exhibited other distin- 

 guishing characters ? 



The objection against the use of the word " hereditary '' in the definition of the 

 term "species" wliich we have here raised is based immediately upon the assumjition 

 of the transmutation of species and their components (varieties). The species A and 

 B, to put it figuratively, may have as sjiecific characters a and b respectively; in the 



* J'apiliii echidna £rom the lesser Sunda Islands is a dwimilis-toTm, but is perhaps specifically distinct ; 

 1 do not take this insect into account here. 



t Collected by Captain Caylcy Webster. The importance of the record of individual aberrations is 

 here again evident. 



