SATURN1A. 299 



rence of hermaphrodites in lepidoptera is given by Speyer as about i in 30000. 

 The latter number he considers rather too low than too high. There was no 

 hermaphroditism in any known members of the families of the progenitors of 

 these hybrids, and its appearance in this proportion must be, he thinks, a con- 

 sequence of their exceptional origin. On the other hand, he has bred more than 

 1000 genuine hybrids without one such case occurring, nor was any such tendency 

 shown by a single one of the sixteen specimens of S. standfussi* — a form whose 

 origin is analogous with that of the present cross-product. In the well-marked 

 hermaphrodite mentioned above, the distribution of sexual characters is remarkable. 

 The shape of the lorewings is rather female than male, the colouring on the upper- 

 side of both is male ; on the underside, the right is mostly male and the left female. 

 In both hindwings, the upper surface has the costal portion male, the remainder 

 female ; the right, which is about one-fifth larger than the left, has the female area 

 more extensive. The under surface of the right or larger hindwing is female, of 

 the left, mostly male. The right antenna is male in form ;.the left partly male and 

 partly female. The external genital organs on the right side are of a malformed 

 male type ; on the left side absent. The want of vigour in this cross is shown by 

 the fact that out of nine pairings, each resulting in an average of 200 eggs, only 

 ten larvae were produced, and of these, only six, as we have seen, attained the 

 perfect condition (loc. cit., pp. 91 — 98; pi. iv., figs. I — 3). 



4. hybr. schlumbergeri, Stdfss. (bornemanni J x pyri ? ). — As S. bornemanni 

 is the cross product of S. pavonia <S X S. spini ? , this hybrid is descended from 

 all three species. Only one pairing was obtained and 92 per cent, of the eggs 

 hatcher, whilst of the resulting progeny a pair is figured (Ent., xxxiii., pi. iv., figs. 1 — 2). 

 The ? gives one the impression of a very large, gigantic S. spini. The $ also inclines 

 mostly to a large S of this type, without, however, entirely discarding its 

 connection with S. pavonia and S. pyri. Two male pupae of this rarity have 

 already gone over three winters without developing and will probably go over 

 a fomth winter without emerging {loc. cit., xxxiii., p. 345). 



Further experiments were carried out by Standfuss in the years 1895, 

 1896, 1897 {Ent., xxxiii., pp. 343^/^.). As has been already noted 

 above, the hybrid males of S. hybr. pavonia x spini and S. pavonia Xpyri 

 can be crossed back with the ?s of either of the parent soecies, and 

 imagines from all these four crossings have been obtained, whilst the 

 three Saturnias — pavonia, pyri and spini — have also been combined into 

 a hybrid form, and still more complicated hybrids obtained, which only 

 reached, however, the larval stage. These are detailed at length in 



tabular form (Entom., xxxiii., p. 344) and figured [loc. cit., pi. in iv 



(vii — viii)]. These crossings gave besides those already enumerated : 



1. S. hybr. dixeyi, Tutt {bornemanni g X spini ? ). — The hybrid figured 

 [Ent., xxxiii., pi. hi., figs. 4 and 5) resembles S. spini in structure and coloration. 

 The <? differs from S. spini, however, in the single coloured dark grey - brown 

 hairs of the upper side of the abdomen, while the ? possesses the broken 

 transverse band at the base of the forewings, which is a peculiarity of S. pavonia. 



2. S. hybr. complexa, Tutt ( standfuss i $ x pavonia ? ). — Of this also very com- 

 plicated hybrid, three males of the same brood were successfully reared. The largest is 

 figured (Ent., xxxiii., pi. iv., fig. 4) the others measure respectively 71mm. and 73mm. 

 The mother and grandmother of these hybrids were both very large ? s of the Dalmatian 

 type of S. pavonia. The great size of these insects is certainly chiefly due to this 

 origin, as only one-eighth of the S. pyri blood of the great-grandmother remains. 

 In general all the specimens have the appearance of large light-coloured S. pavonia 



c? s of the Dalmatian form. 



Three other crossings were obtained, fertile ova and larva? result- 

 ing, but the larvae died of an infectious disease when nearly fullgrown. 

 These were: (1) S. hybr. pavonia $ X schaufnssi $. (2) S. hybr. 

 schaufussi$ x pavonia °. ■ (3) S. hybr. bornemanni $ X schazifussi $ 



It might be supposed from these results that these crossings 

 were abundantly fertile inter se, but this is not so, for Standfuss has 

 found the $ s almost entirely incapable of egg-production, and has 

 discovered as yet in the g s only a very few degenerat e egg-germs or 



* Obtained in some numbers later, seepostea, pp. 302—304. 



