04 Riley — Sotiic Inter rclalio its of rUtid.'^ (ind Jiusecls 



ways. The })lant never produces seed where Pronul)a does not 

 <ixist ; it never produces seed when she is excluded artificially, 

 and experiments which I have made with artificial or" brush 

 {)ollination all show that it is much more difficult to ensure 

 complete fructification than would at first api>ear, and that the 

 act of pollination is rareh^ performed with a l)rush or by using 

 the fiower's own filaments, as successfully as it is done by Pro- 

 nuba. It is Promiha y n cxaaella whicli pollinizes all our Yuccas 

 east of the Rocky Mountains, so far as known, and the S])ecies 

 is remarkably uniform in character, its appearance in time being 

 coetaneous with the fiowering of Yuccri JiJamentosa. On the 

 western plains its appearance has become adaj^ted to the flower- 



Fui. !). — Mature puds of Vucca anyusiifolia : a, artitiinally pulliuized and protected 

 from ProiiLiba; 6, normal pod, showing eonstrietions resulting from Pronuba punctui'e 

 and exit lioles of larva : c, one of the lobes cut open, showing larva within. 



ing of Yucca anyuMtifolut., but in the east, where these two species 

 of Yucca are frequently grown side by side, Y. aiu/ustifoUa lowers 

 two or three weeks earlier than Y. filavientosa and generally too 

 early to receive the visits of Pronuba, so that it i)roduces seed 

 only on very rare occasions. Yticcd hrevlfvlia is pollinized by 

 Proiiiiha synthetlcii Riley, the most remarkable species of the 

 genus, having very stout maxillary tentacles, a very stout ovi- 

 positor, shorter than that of yticcascJId, Init characterized chiefly 

 by having fuliginous and unsealed wings and a polished, naked 

 and flattened body — structures all well adaj)ted for crawling be- 

 tween and about the comi)act and crowded flowers, with their 



