xvi 
that they may be supposed to pass into Iridacez through Thelymitra and 
Sisyrinchium on the one hand, and into Zingiberacez through Phrynium 
and such a genus as Evelyna * on the other. 
With regard to Apostasiacee, their relation to that order does not appear 
to me to be greater than to either of the two now mentioned ; and in the 
absence of all evidence as to the connecting links which join Orchidacez 
and Apostasiacez it seems unnecessary to advert further to the subject. It 
may, however, be observed that Apostasia has apparently as much claim 
to be regarded as a diandrous monadelphous Amaryllidaceous plant allied 
to Hypoxide and standing, perhaps, in the same relation to that order as 
Gilliesia to Liliacez, as it has to be regarded as a trilocular Orchidaceous 
plant with the gynandrous organization lost. 
In classifying this order the most important characters appear to reside 
in the pollen, which in many is consolidated into firm waxy masses of a 
definite number in each species, and in others is either in its usual /oose 
powdery condition, or is collected in granules or small wedges the number of 
which is far too great to be counted. 
Of those with waxy pollen masses some (MaraxEz) are destitute of 
any visible processes by which the masses are brought into contact with 
the stigma ; others (ErrpENDREx) have strap-shaped caudicule, which are 
either bent down upon the masses themselves, or serve to hold them 
together, without, however, forming any organized union with the stigma ; 
whiie the remainder (VaxpEx) have a caudicula, which adheres firmly to a 
gland found on the upper margin of the stigma and separating freely from 
that organ. The last form is much more distinct from the two first than 
they are from each other, and I am inclined to think that hereafter it will 
either be requisite to combine Malaxee with Epidendrez, or to exclude 
from the former not only Acanthophippium, Celogyae and Pbolidota, but 
several other genera at present referred to them. 
The genera with powdery, granular, or sectile pollen cannot be classified 
so conveniently by modifications of that part, but are readily divided into 
3 natural tribes by peculiarities in the anther. In some (Ornunrsz) the 
anther is erect, not hinged to the column but continuous with it, and 
stands above the stigma, the pollen masses having their points directed to 
the base of the lobes of the anther ; in others (AnETRHUsEX) the anther is 
hinged to the column, upon the end of which it is placed transversely like 
alid; and in others (NEorrxx) it is also hinged to the column, but is 
stationed at its back so as to be nearly parallel with the stigmatic surface. 
If to this we add that Cypripepex have two anthers, while all the 
others have one only, we find the order divided into seven tribes, of which 
the following is a tabular view. 
* This genus has been 
present work was 
The Ble 
en published in Póppig and Endlicher's genera plantarum while the 
y vas passing through the press and will be introduced into the “ Revision," 
tia capitata of R, Brown, p. 123, is a species of it. 
