Pairing of the Plebeiid Blue Butterflies. 169 



This is so, as far as I have observed, though I have not 

 prepared long series of specimens in more than two or 

 three species. 



The extraordinary extensibility afforded by the prop 

 and rein does not occur, I think, elsewhere ; I can, however, 

 rely only on my own small knowledge of these structures, 

 which may easily be at fault. Some other groups of the 

 Lycaeninae show what we may regard as initial stages 

 of the Plebeiid structure. For example, Lampides boeticus 

 (PI. LV, LVI), which has an immense terminal armature 

 to the ostium extremely unlike anything in the Plebeiids, 

 nevertheless retracts and extends this mass for a distance 

 equal to its own length, as is seen by comparing the photo- 

 graphs on PI. LVI. Iolas (PI. LIII), again, on the other 

 hand, with an equally large armature, seems to have it 

 much less movable. L. argiolus (PI. LIV) also has the 

 armature slightly movable, but apparently not retractile. 



L. alcon and euphemus (PI. LVI I) seem to be retractile. 

 T. telieanus and E. argiades (PI. XL VIII) are also apparently 

 retractile, but none of these to more than a small amount. 



I have put on the same plate as the two last (PI. XLVIII) 

 a photograph of P. martini ; this is quite a typical Plebeiid, 

 but in the specimen, the parts are only partially exserted, 

 the prop perhaps one-third extended, and the rein there- 

 fore equally enclosed. In this species the rein is chitinised 

 near its extremity, as in the species shown on PI. LIV, 

 in zephyrus (PI. XLIV), in meleager and admetus (PI. 

 XLI), in pylaon and candalus (PI. XXXVI). The result 

 is to give the two specimens in PI. XLVIII— martini, 

 fig. 2, and argiades, fig. 3 — the appearance of being of very 

 closely related structure, which is not really the case. 



With this reference to chitinisation of the extremity of 

 the rein, it may be as well to include the circumstance that 

 in other species there is a chitinisation towards its base. 

 This occurs in C. semiargus (PI. XLVI), A. isaurica and 

 donzelii (PI. XLV), A. escheri (PI. XL), and to a trifling 

 extent in some others, in which it may or may not be 

 really part of the loop of the prop. 



I have frequently observed these structures in the field, 

 when functionally active, and have for several years made 

 efforts to secure specimens preserved for observation and 

 examination in that state, but have always failed, though 

 on several occasions I felt sure up to the last moment 

 that I had succeeded. 



