MJt, U. EENXnAM OX MrHTACE.!:. 123 



deeply uotclied at tlie insertion of the radicle. In xno^t Eucali/pti 

 this notch is so deep as completely to divide each cotyledon into 

 two broad conduplicate plates enveloping the radicle. In the 

 large and natural genus Melaleuca there are species "with narrow 

 and with broad cotyledons, flat or variously folded, very frequently 

 the two cotvledons so folded that each one embraces one half of 

 the other. 



The tribe of Myrteae proper, after abstracting a few anomalous 

 monotypic or very small genera, comprises, according to the lowest 

 estimate, above a thousand species, which some botanists have 

 multiplied to nearly twice that number, all with so little diversity 

 of floral structure and habit that, without the aid of the embryo, 

 they might well have been included in one natural genus. The 

 embryo, however, presents three remarkable types, which, wben 

 first observed, were supposed not only to be widely and constantly 

 distinct, but also to correspond with differences in the number 

 of parts of the flower, in the texture of the testa, and other minor 

 characters, and they were gladly seized upon as absolute tests of 

 three great genera or subtribes. These are : — 1, the horseshoe- 

 shaped, circular, or spiral embryo o£ Jfi/rfns, PsiJium^ and their 

 allies, consisting of a long terete radicle, with two very small 

 cotyledons at the inner end ; 2, the broad, thin, very much folded 

 or contortuplicated cotyledons oi Myrcia, more or less surrounded 

 by the curved terete radicle ; and, 3, the thick fleshy embryo of 

 Eygenia^ sometimes apparently homogeneous, but more frequently 

 showing the line of separation of two thick hemispherical coty- 

 ledons, connected by a very short radicle. Eelying upon these 

 characters maintaining throughout the constancy observed in the 

 Tery few species which he could examine, DcCandolle was enabled 

 to make an apparently excellent distribution of the great mass of 

 succulent-fruited Myrtea? into the three principal groups above 

 mentioned. It was further believed that, as between the three 

 typical genera, Hugenia^ ATijrtus, and Myrcla^ the former might 

 always be known by its 4-merous flowers, and that the latter two, 

 with 5-mcrous flowers, might be distinguished by their inflores- 

 cence and the number of ovules. Further observation has not, 

 howevei", confirmed this neat demarcation. A large number of 

 South American, especially Chilian species which, from their 

 4-merous flowers, had been placed in Eifgenia, prove to have the 

 embryo oi ITyrtus \ the cymose, Avell-developed inflorescence some- 

 times accompanies the embryo of Myrtus as well as that oiMyrcia ; 

 and a few instances are known of embryos intermediate in form 



