in Myzodendron punctulatum , Banks et Sol: 187 
matter. The plumose appearance is caused by the separa- 
tion of some of the utricles which diverge on all sides in the 
species in which the setae are terete or in their opposite 
margins when the latter are compressed. 5 
My own observation has led me to believe the setae to be 
very definite structures which must be given an important 
place in a description of the morphology of the flower. 
Entering each seta at its base, and reaching to near its free 
apex, is a single central vascular bundle of which the pro- 
toxylem is the more marked part (Figs. 4 and 8). The rest 
of the seta consists of parenchymatous tissue, covered by a 
definite meristematic epidermis, from which at irregular in- 
tervals long upright hairs (trichomes) grow out, and stand 
parallel to one another, with their free tips reaching to the 
top of the groove in which the seta occurs. Small and hidden 
in the flower (Fig. 4) the seta in the mature fruit is very 
conspicuous, and gives the fruit a characteristic appearance 
(Figs. 1, 9, &), A longitudinal section of the elongated seta 
shows the elements of the vascular bundle much drawn out, 
especially in the upper part of the seta. I made a number 
of experiments to test the presence of viscidity in the seta 
and its hairs. The free ends of the hairs (Fig. 10) were the 
only parts which, before treatment, appeared to be viscid, 
but in no part of the seta or its hairs was I able to find 
viscidity, a conclusion reluctantly come to and one with which 
I am not satisfied. The experiments I am about to detail 
were accompanied by control ones on fresh viscid matter. 
As the material of Myzodendron had been in spirit since 1882 
the more important of the control experiments were also made 
on the mucilage sacs of the roots of Angiopteris , material of 
which had been in spirit as long as Myzodendron . The possi- 
bility that the absence of a trace of viscidity in the Myzo~ 
dendron seta might be due to the action of the spirit was 
found, judging from Angiopteris , to be groundless. Each 
seta-hair is an independent cell, with a nucleus and very 
much cell-sap, cut off from the epidermis and not like a 
root-hair, an outgrown epidermal cell. The whole hair is 
