144 GORDONIA. 



mucilage glands, whose function is to keep the young tissues from 

 injury while still packed in the bud. When the leaves expand, 

 they are already dead in the Asiatic species but I have seen them 

 to be still alive on young expanded leaves of the American 67. pubes- 

 cens, and perhaps their structure differs. It is interesting to link 

 this protective function with the demand that the tree makes for 

 moist air. A little nerve runs out to the glands. This same type 

 of gland occurs in allied genera such as Adinandra, as in Gordonia 

 imbricata and G. Scortecliinii, where the leaves are toothless. The 

 apex to the leaf also is glandular, and dies as or before the leaf un- 

 folds, often leaving an emargination. Some of the earlier botanists 

 have tried to make a character of this emargination, but the degree 

 to which it has gone in any particular leaf may be but a response 

 to local or to temporary air conditions. Eeference may be made 

 to fig. 15, on p. 158, where two flushes are represented. At the 

 commencement of the first, weather conditions appear to have been 

 such as to destroy with the apical gland quite a distinct piece of 

 the neighbouring tissues : these conditions were not repeated. Hairs 

 are present on the young parts which vanish like the glands more 

 or less completely with age: in the leaf bud they are protective 

 similarly. They afford no good specific characters. 



The Flowers. 



Flowers are formed on very short axillary shoots in the upper 

 leaf axils of a flush or in the axils of cataphyllary leaves following. 

 If the terminal bud should die they appear as if terminal on the 

 foliar branches, and this seems particularly; to happen in G. 

 penangensis. In this species the weight of a falsely terminal 

 capsule on the rather slender branchlets makes it to hang; but in 

 most species they dehisce upwards, see fig. 7 of G. hirtella. 



The Gordonias, as said, may be got to flower at relatively 

 small sizes. -No information is available as to subsequent seed- 

 formation; but in Singapore 67. singaporiana already seeds at a 

 height of 6 meters. 



~No information is available as to the formation of flowers in 

 Gordonia sexually imperfect, except that a female condition was 

 found in the first observed plant of 67. axillaris. However as 

 Urban (Bericlite d. deutsch. hot. G es ells cli aft, xiv, 1896, p. 51) 

 has found male and female flowers in the Tropical American 

 species of Haemocliaris, the occurrence of sexually imperfect flowers 

 in a Gordonia is hardly isolated. Further Urban ascribes to the 

 male flower a short style, which makes the definition between Gor- 

 donia and Ilaemocharis just nothing. 



The old trees of G. singaporiana in the Botanic Gardens, 

 Singapore, have their flowers fully hermaphrodite. They open in 

 the evening or after dark, facing horizontally or slightly down- 

 wards, and fall during the next forenoon. They have a smell which 

 is fairly strong, but hardly fragrant; they are in colour creanry 

 white : they are produced over many months of the year. Infertile 



Jour. Straits Branch 



