266 Sargent . — Notes on the Life-History of Pterostylis . 
The processes on the internode next below the surface terminate in long 
thread-like cells, which give access to the mycelium of a fungus. This 
gradually penetrates to the cortex, filling its cells with a tangle of hyphae. 
I only discovered the presence of this fungus last year, and know very little 
concerning it. There is no doubt, however, that the orchids depend very 
largely on their fungal partners for a living. Their root system is always 
very scanty, only a few short roots being developed from the lowermost 
internode. 
When the plant has spread its leaves and gained sufficient strength, 
a shoot is developed just above the tuber. This shoot grows straight down 
and ultimately thickens into a tuber, which replaces the old one, and forms 
the resting stage of the plant during the next summer. In the species 
reflexa , , constricta , nana^ and pyramidalis , secondary tubers are also formed 
in the following manner : one or more branches appear just below the fungal 
tissue. These grow horizontally, or slightly downwards until they attain 
a length of two or three inches. They then turn abruptly down and grow 
a little longer. The turned-down parts then thicken and become tubers. 
In this manner two or more plants are formed from the original one. It is 
doubtless, chiefly owing to this vegetative mode of reproduction, that these 
species are usually found in colonies, in which the plants are often very 
much crowded. I am sure, however, that seedlings assist in the formation 
of the colonies, as I have repeatedly found them therein. The other Westra- 
lian species do not form colonies, and I have never found them with long 
tuber-forming branches. Occasionally, however, they form two tubers 
instead of only one beside the old tuber. Usually only one of these pro- 
duces a flowering scape the first year after formation. 
The leaves of the young and non-flowering plants of all the Western 
Australian species are closely similar. Speaking broadly, they are ovate in 
outline, are borne on amplexicaul petioles (usually short), and are arranged 
in radical rosettes. P. rnfa and nana retain this form of foliage all through 
their lives. The flowering plants in these species throw up scapes through 
the centres of the rosettes. In P. pyramidalis , turfosa , and barbata the 
lower internodes elongate at flowering time, so that the rosettes, as such, 
vanish, though the leaves still remain near the bases of the stems. In the 
remaining species the flowering and non-flowering plants are sharply differ- 
entiated. The latter, as I have stated, produce rosettes of ovate leaves ; but 
when a tuber of one of these species is in a fit condition to flower it throws 
up a strong scape bearing at intervals alternate, sessile, lanceolate leaves. 
There is no rosette at any time of the year in which the plant flowers. 
The scape varies in height from a few inches ( P . nana , , reflex #, &c.) to 
a foot or more ( vittata and recurva). The inflorescence is a loose few- or 
many-flowered raceme. In some species (; nana , pyramidalis , reflexa , and 
constricta ) it is reduced to a single flower, with usually the rudiment of 
