522 EVOLUTION OF THE EUCALYPTS, 



bably offshoots from the Stringybarks. Such forms of " entire " 

 cotyledon as we see in E . dumosa, E. incrassata, E. populifolia, 

 and E. qiiadrangulata, are either reductions, or, more probably, 

 due to a return to the ancestral type after reduction through an 

 emarginate ancestor. In the emarginate type of reduction, we 

 see what the Eucalypts have found the most suitable, and so 

 given the preference to. They evidently found it better to cut 

 out and reduce, than to reduce only. In E. inarginata, we see 

 how the size, still great, has been slightly reduced by the 

 emargination. In E. Planchoniana, the same obtains, and this 

 species, with its parallel anthers, but an oil and leaves like those 

 of E . pUularis, probably represents most closely the ancestral 

 prototype of the group to which E. pilularis, with its reniform 

 anthers and smaller cotyledons, belongs. I have represented 

 diagraramatically in fig.D of Plate Ixix., the reduction from the 

 cotyledon of E. corymhosa to that of ^. ehvophora oi the ylohulus- 

 type. It will be seen that the lightly shaded, apical portion has 

 been entirely removed, but the lobed part still remains. When 

 one takes away, in addition, the deeply shaded lobed portion and 

 the remainder of the part at the bottom of the notch, we have a 

 cotyledon of the type of E . cornuta, leading up to E. squamosa. 

 The narrowing of the angle and reduction of the size of the limbs 

 have led to the extreme form of Y-shaped cotyledon seen in E. 

 gracilis and E . pendula. From the predominance of such forms 

 in Western Australia, it seems likely they arose there in the first 

 instance, and gradually spread east along the Australian Bight. 

 Hence we find, in the far west of New South Wales, and in 

 Victoria and South Australia, such species as E. gracilis, E. 

 cneorifolia, E . pendula, E. unciriata, E. calycogona, and E. 

 oleosa. The most interesting of all of this type is E. squamosa. 

 This occurs near Sydney on the Hawkesbury Sandstone, and also 

 near Rylstone on the western slopes of the Dividing Range. It 

 appears to be a retrograding species, as the moist conditions do 

 not seem to be suitable; and everything points to its having 

 migrated from the hot, dry west, and crossed the Dividing Range. 

 The longer cotyledon-limbs indicate a response to the cooler, 

 moister conditions. In the three subdivisions of the great 



