Flowering Plants of New Zealand. 97 



make the plaut practically mouoecious. Again, in some 

 polygamo-dioecious plants as Fuchsia, Aristotclia, Leptos- 

 permum, &c., truly hermaphrodite forms occur; while in 

 others as Clematis^ Pimelea, &c., the flowers are struc- 

 turally hermaphrodite, but owing to imperfection of one 

 of the sexual whorls, functionally dioecious. With regard 

 to the Comjjosita:,, which I have included under the head 

 polygamo-dioecious, most of the flower-heads, as is usual 

 in the order, include both hermaphrodite and pistillate 

 flowers, but in the majority of cases, the former are so 

 distinctly proterandrous, as to be unisexual in function. 

 The following is a summary of my observations : — Of the 

 262 species referred to, 129 were invariably herma- 

 phrodite, but of these 37 species were so strictly proter- 

 androus, and 10 species so strictly proterogynous, as to be 

 functionally dioecious ; 22 species were invariably dioeci- 

 ous (including 14 species of Cojorosina), and no less than 

 79 species polygamo-dioecious (including 31 species of 

 Compositce). If then, from the total number observed, 

 we subtract all which are more or less unisexual either 

 structurally or functionally, we shall find that only 82, or 

 a little over 31 per cent., have any chance of being self- 

 fertilised. But even of those, many are so specially con- 

 trived, such as the majority of the Orchidem, that even they 

 must be excluded from tlie list. 



Again, out of this total of 262 species, 139 have con- 

 spicuous flowers, 83 being conspicuous alone, and 56 by 

 association. These are roughly subdivided as follows : — 

 white, 72 species; yellow, 27 species; greenish, 18 species; 

 red or pink, 11 species ; purple or lilac, 9 species ; and 

 blue, 2 species. The remaining 123 species are more or 

 less inconspicuous, but even among them are to be found 

 many which are absolutely entomophilous, and depend (as 

 in Tupcia) on their fragrance and honey to attract insects. 



The total number of species observed to secrete honey 

 was 99, but it is to be remembered that flowers do not 

 contain the necjtar at all periods of their development, but 

 that it is produced only at such times as the visits of 

 insects would be of service. I have frequently noticed and 

 recorded that certain flowers produce no honey, and then 

 have come upon these same flowers on a warm, bright day, 



TRANS. BOT. SOC. VOL. XIV. G 



