KNOWLEDGE OF MONOCOTYLEDONOUS SAPROPHYTES. 211 
vessels. There is no mechanical tissue, and the walls of all the 
cells are thin and unmodified cellulose (excepting xylem and 
feeble cuticle). 
No stomata occur in Epipogum nutans or in Galeola javanica. 
Vascular Bundles of Holosaprophytes. 
Hitherto in holosaprophytic monocotyledons a normal com- 
plete vascular bundle has been described by Poulsen alone as 
occurring (in Sciaphila). Statements as to the simplicity of the 
bundles are contained in the works of Johow and Reinke. And, 
as far as I know, amongst dicotyledonous holosaprophytes, only 
F. W. Oliver (8) has given an account of a complete typical 
vascular bundle (in Sarcodes). 
Yet in the holosaprophytic orchids on which I have worked 
there is in the vascular bundles always a clear distinction into 
aylem and phloém. As regards the xylem, it is quite typical in 
the floral axis, but is more feebly developed in the roots, where 
no spiral or annular vessels occur. There appears to be a 
tendency to form tracheides with transverse narrow pits, in place 
of forming vessels with round pits or other modes of thickening. 
This is an interesting point, in that the wood thus displays a 
structure more primitive in both an ontogenetic and a phylogenetic 
sense. Inall cases the phloém is highly differentiated throughout 
the plant (except in the finer bundles of the scales), and possesses 
sieve-tubes, companion cells, and parenchyma. The general 
peculiarity is that the parenchyma is more developed, and the 
sieve-tubes are smaller and fewer than in atypical bundle. This, 
again, is a more primitive condition. 
Degeneration in the Shoot. 
As a saprophyte gives up its autotrophic mode of nutrition, it 
is the leaves which in particular lose their great importance as 
organs of nutrition. And we shall expect that they will display 
more complete phenomena of degeneration and change than any 
other parts of the plant. Such, in fact, is the case. The leaves 
of holosaprophytes are reduced to mere simple scales. 
It has been noted by Darwin that when an organ is under- 
going atrophy, it is liable to vary through much wider limits than 
is the case with an ordinary organ. If, then, considerable 
variation be of any assistance in hastening the evolution of an 
organ, we might anticipate that a degenerating organ would 
