546 DEVELOPMENT OF THE N.O. MYRTACEjE, 



tus, originated in Australia, and then, in some way, crossed the 

 oceans to Asia, Africa, and America, giving rise there to types 

 very similar to each other; nevertheless differing widely from the 

 capsular types? Or are we to consider that the localised form is 

 the more or less depauperate descendant of a type belonging to the 

 moist and fertile tropics'? 



An application of the Law of Probabilities suggests that either 

 the localised and isolated xerophytic, or the more or less depauper- 

 ate forms, which all exhibit great morphological differences among 

 themselves, are the descendants of types, possessing at once a 

 range world-wide as regards the fertile tropics, a close similarity 

 of morphological characters, and a great wealth of species. This 

 idea is strengthened by the fact, that the Cretaceous Period was 

 one of genial and moist climate, and that Australia has been 

 isolated from the tropical world since the close of the Cretaceous. 



A consideration of these points suggests an appearance as out- 

 lined hereunder for the earlier forms of the family: — A tree or 

 shrub, generally large. Leaves simple, entire, opposite, penni- 

 veined, dotted and possessing intramarginal veins. Calyx-lobes 

 and petals imbricate. Petals 4 or 5 (probably five). Flowers regu- 

 lar, solitary or in cymes. Stamens indefinite, numerous, free, often 

 crimson or brightly coloured, and generally exceeding the petals 

 Anthers two-celled, versatile, the cells parallel and opening in 

 longitudinal slits. Ovary inferior with two or more cells. Style 

 simple. Ovules two or more to each placenta. Fruit inferior and 

 crowned by the persistent limb, indehiscent, succulent or fleshy, 

 rarely dry. Albumen none. Cotyledons thick and fleshy, with a 

 short radicle. 



This plant appears to have possessed a graceful and umbrageous 

 habit, with dense, glossy and beautiful foliage. Wherever possible, 

 exposed situations were avoided, as also sandy porous soils. 



Home of the Earlier Types. 



This difficult question is best approached from three view-points, 

 namely, the areas of greatest density, as regards species, at the 

 present time; the location of modern types most closely allied to 



