4 o6 AN GIO SPERM AE DICOTYLEDON ES 



three others take their place. The pollen continues to be shed for about 

 days, and during this time the styles with their immature stigmas lie close toget 

 It is only after the anthers have emptied themselves that the styles elongate 

 diverge, so that the stigmas occupy the position taken up by the anthers in 

 first (male) stage. 



Kirchner says that the flowers vary in size without any other correlated 

 differences. 



Visitors. I have seen a hover-fly (Eristalis arbustorum Z.), skg., and also 

 a beetle (Meligethes) at Kiel. Sprengel describes pollination by the blow-fly 

 Calliphora vomitoria L. 



Herm. Miiller (H. M.) and Buddeberg (Budd.) observed the following. 



A. Coleoptera. (a) Curculionidae : i. Miarus graminis Schonh. (b) Der- 

 mesiidat: 2. Anthrenus scrophulariae L. (H. M.). B. Diptera. (a) Empidae: 3. 

 Empis tessellata F., skg. (H. M.). (b) Syrphidae: 4. Eristalis arbustorum L., skg. 

 (H. M.). C. Hymenoptera. (a) Apidae : 5. Andrena schrankella Nyl. J, skg. 

 (H. M.) ; 6. Halictus malachurus K. 5, skg. and po-cltg. (H. M.); 7. H. minutissimus 

 K. $, do. (H. M.) ; 8. H. morio Z. $, do. (H. M.) ; 9. H. nitidiusculus K. % do. 

 (H. M.). (b) Tenthredinidae : 10. Cephus sp. (H. M.). 



MacLeod noticed 2 short-tongued bees and 4 Muscids in the Pyrenees (Bot. 

 Jaarb. Dodonaea, Ghent, iii, 1891, pp. 423-4). 



964. S. tridactylites L. (Sprengel, 'Entd. Geh.,' pp. 244-6; Herm. Miiller, 

 ' Weit. Beob.,' I, p. 297.) Sprengel says that the flower mechanism of this species 

 agrees entirely with that of S. granulata. He also quotes the observation of Linnaeus, 

 'sub florescentia germen stylo stigmatibusque destitutum,' which cannot be in- 

 terpreted otherwise than as meaning that the styles and stigmas do not mature 

 until the anthers have withered. 



Hermann Miiller, on the other hand, describes the small white flowers as feebly 

 protogynous. When they open the stigmas are already mature. Soon afterwards 

 the anthers dehisce successively, first those of the outer whorl, then those of the 

 inner. As this takes place they regularly come into contact with the stigmas, 

 effecting automatic self-pollination at an early stage, and this is completely effective. 

 In dull weather the flowers remain shut, or close if they had previously opened. 

 During such weather the nectary, which surrounds the style as a yellow fleshy 

 ring, does not secrete, but in the sunshine it produces glistening drops in the 

 midday hours. 



Kerner states that, besides hermaphrodite blossoms, there are pseudo-herma- 

 phrodite pollen flowers and fruiting flowers on the same plant. Moreover, Warnstorf 

 also says that the protogynous hermaphrodite flowers are associated on the same 

 stock with pseudo-hermaphrodite ones of both kinds ; sometimes the carpels aborting, 

 sometimes the stamens. 



965. S. tricuspidata Rottb. (Warming, Bot. Tids., Kjobenhavn, xvi, 1888, 

 pp. 22-5.) Warming states that the stellate flowers of this species are at first 

 slightly protandrous in Greenland, soon becoming homogamous. Self-pollination 

 is therefore always possible, though the probability of its occurrence varies in different 

 cases. As ripe fruits have been observed, it must be effective. The nectar is 

 exposed. Besides hermaphrodite flowers, Warming noticed purely female ones. 



