330 VARIABLE DICECISM IN PITTOSPORUM UNDULATUM, 



Again, in 1902, Mr. R. H. Cambage* brought evidence forward 

 confirmatory of the observations made by tlie previously quoted 

 botanists. During the last few years I have had opportunities of 

 making close observations on several trees of P. undulattbm grow- 

 ing in my garden, at Petersham, near Sydney, and on numerous 

 others in gardens in the vicinity, which have yielded further 

 information of an interesting nature. 



Usually towards the end of June, in the neighbourhood of 

 Sydney, the flower-buds on the male or staminiferous trees are 

 well advanced, many of them being on the point of opening, while 

 on the female or non staminiferous trees nothing but leaf-buds 

 are visible. Tlie blossoms on the latter begin to open about three 

 weeks or a month later than the others, and are accompanied by 

 a succession of staminiferous flowers on the adjacent male trees 

 during the whole period of flowering. The male trees are much 

 handsomer than tlie female, because of the larger size of the 

 blossoms and the bright yellow of the anthers; while both are 

 fragrant, secrete nectar and are freely visited by bees. 



I have repeatedly noticed a few seed-vessels occurring, either 

 solitary or in clusters, on the male trees, and have ascertained by 

 trial that the seed contained in these is fertile and germinates as 

 readily as that from the female trees. In normal blossoms, when 

 fully open, the ovary of the staminiferous type, though of about 

 equal length to that in the female flower, is not nearly so stout 

 and globose; while the stamens are long, and have the anthers 

 projecting above the top of the pistil (PI. ix., figs. 2-3). In the 

 female flowers, as has been very clearly described by Mr. Hamilton, 

 the stamens are mere rudimentary scales appi-essed to the base of 

 the ovary(Pl.ix., fig.5). With the aid of my sons, I have, in 

 several successive years, made a close examination of many 

 hundreds of blossoms on both kinds of tree; and have found, on 

 the male tree, blossoms having short stamens, with shrivelled 

 abortive anthers which did not dehisce or form pollen. By 

 marking these with little pieces of cord tied round the petioles, I 

 was able to keep them under observation, and found that these 



* Proc. Linn. Soc. N. S. Wales, 1902, p.593. 



