THE president's ADDRESS. 27 



among 450 genera, and occurring almost everywhere where con- 

 ditions render possible the existence of flowering plants except 

 in the coldest regions. The Orchids supply one of the best 

 examples of remarkable variety of detail in adaptation of a type 

 of flower, formed on a comparatively simple plan, to the effecting 

 •of pollination by means of insects. In this case the reason for 

 the adaptation is obvious ; we know but little as to the details 

 of pollination of exotic species of orchids, but may assume that 

 the amazing variety of form and colour is associated with a 

 correlated variety in the type of visiting insect. 



But there is another great and even more widely spread family 

 in which the form of the flower, coupled with the form of the 

 inflorescence, shows a variety comparable with that of the 

 Orchid. I mean the grasses. What has been the determining 

 factor here ? Grasses are wind- pollinated, and the breeze is a 

 blind agent. The pendulous anthers and protruding stigmas 

 •are eminently adapted for this purpose, but there is an infinite 

 variety in the form and detailed structure of the outer covering 

 or " glumes " which can have no relation to wind-pollination. 

 In some cases the distribution of the ripe seed is aided by the 

 form of the glume or its appendages, as, for instance, by the 

 presence of awns or stiff or silky hairs ; but this factor will not 

 explain the great majority of the variations in details of structure. 

 In many cases the parts concerned have fallen before the grain 

 is dispersed, or have lost so much of their characteristic form 

 that the species cannot be satisfactorily determined. Are we 

 dealing here with merely indiscriminate variation ? The dis- 

 ■crimination of species in some genera is notoriously difficult, as for 

 instance in the Poas and Fescues, and the explanation may be that 

 there is nothing to check variation. Indiscriminate variation 

 would be checked at once in a type of flower so nicely adjusted 

 to pollination by a particular insect as are many orchids — it 

 would be suicidal. But in a wind- pollinated flower a wide 

 variation might have no effect on the transmission of pollen. 



An interesting case of secondary adaptation is found in the 

 so-called viviparous grasses, which occur especially in high 

 latitudes and upon high mountains where ripening of fruit is 

 often uncertain. Entire spikelets or single flowers, with floral 

 glume and pale, are transformed into small-leaved shoots which 

 Tbear at the base the beginnings of roots. These miniature 



