374 THE CANADIAN ENTOMOLOGIST. 



Bombus, Xylocopa, Ceratina, Eugiossa, Ce?it7-is, and other genera. These 



collect only pollen, since nectar is wanting. Of forms other \.h3.n SolaJiutn 



and Cassia we know little. One species of Dichorisandra examined in 



the living condition seems to show no secretion of free nectar. Several 



members of the family are known to be adapted to pollen-collecting 



Apidce, and it is probable that the apically dehiscent forms are too. 



Analogy leads us to the same conclusion for Monochoria of the Pontede- 



riacefe, and for representatives of the Liliaceae, Amaryllidacese, Pittos- 



phoraceae, Tremandraceie, Ochnacese, Dipterocarpaceee, Myrsinaceae, 



Loganiaceae and Rubiacese, while for Ouraiea of the Ochnacese, Begojiia 

 of the Begoniacese and Exacum of the Gentianacese we have observations 

 which indicate the validity of such an assumption. In this place space 

 cannot be devoted to the structural and ecological comparisons which 

 evidence strongly in favor of regarding all these forms as adapted to 

 pollination by the larger pollen-collecting bees. 



Our direct knowledge of the ecology of members of the Mela-stoma- 

 taceous type is not extensive, and some of the more detailed studies have 

 been made on forms which are anomalous rather than typical. It may 

 be stated with much confidence, however, that the type is primarily as 

 truly adapted to the larger Apidee as is the Solanum-Cassia type. 



We may now turn our attention to the only phase of the problem to 

 be especially considered here. 



It is evident that if certain plants are closely dependent upon special 

 groups of insects for their pollination they will be limited in their 

 geographical range by the distribution of these insects. This is exactly 

 the condition which we seem to have in the present case. As has just 

 been suggested, these forms are apparently adapted to pollination by the 

 x^pidse, and the Apidae are represented in the faunas of the several main 

 divisions of the earth's surface in very different numbers, and this seems to 

 determine the proportionate representation of these apically dehiscent 

 types. Our knowledge of the Apidse and their geographical distribution 

 is still very incomplete, but such data as we have indicate that the bees 

 occur in much greater abundance in tropical and extra-tropical South 

 America, the Indian and the Australian regions than in many other por- 

 tions of the globe. Because of the incompleteness of our data concerning 

 the insects, the geographical distribution of the plants will be taken up 

 first. 



