ON THE GERMINATION OF ORCHIDS, 293 
lumps similar to those which they form in the cells of the roots. 
Certainly these fungi in the natural course could live freely in the 
_soil and propagate themselves from neighbour to neighbour; but they 
never produce spores, for they appear to be entirely unable to produce 
germs which can be disseminated in the air, a property so common in all 
other moulds. 
I know three species of these orchid fungi; they belong to the Rhizo- 
ctonia genus; I have not yet given them any names; but a simple look 
at culture tubes in which they are confined, like those in fig. 91, enables 
one to distinguish them easily. One species seen indifferently in the roots 
of Cypripedium, Cattleya, Cymbidium, and Aérides appears to- be the 
most widely distributed. A second species has not yet been met with 
outside the roots of Vanda and Phalenopsis. A third I have only found 
in the roots of an Odontoglossum. Other species of these orchid fungi 
doubtless exist, but in relatively limited numbers, assuredly much fewer 
than the number of orchid species. 
Although the fungi can live apart from their host-plants, the orchids 
themselves require the presence of their guests for their own development. 
I have sown the seeds of numerous orchids “aseptically ’’ in sterilised 
tubes, on cotton or in jelly, saturated with a nutritive solution. I have 
for many months protected the seeds under covers from micro-organisms, 
in suitable temperature and light. In these conditions the seeds have 
not freely germinated ; they swell, and later on they get green, but their 
growth remains insignificant. 
On the other hand, if germs of the appropriate fungus are sown 
with the seeds, they commence to germinate almost immediately in a very 
regular manner. With the fungi which I possess, I have obtained not 
only the germination of Cattleya and of Cypripedium, but also of 
Odontoglossum, Phalenopsis, and Vanda, which are regarded by practical 
men as more difficult. 
Fig. 91 shows culture tubes in which I have raised some of these 
plants, and also a tube in which the seeds haye been placed without fungi, 
and are hardly developed at all. 
In nature or in the greenhouse the same thing takes place as in my 
culture tubes. I have examined a large number of young orchids which 
had germinated in very varying conditions, and I always noticed that 
they were invaded by the fungus from the beginning of their life. The 
orchids are therefore practically dependent on their parasitic fungi, since 
they do not grow without them. 
One of the greatest difficulties in conducting such experiments is to 
procure the fungus suitable for each kind of orchid seed. As a rule, I 
use fungi collected from the roots of an orchid of the same species, or 
of one nearly allied to that which bears the seeds. This is the best rule 
I can give, but it may not be an absolutely sure one. The properties of 
the fungi may indeed be different at the time of gathering from what 
they are at the time of using them. Also the fungi of Cypripediwm and 
of Cattleya, kept cultivated in my tubes for three years, have lost the 
ability to cause the corresponding seeds to germinate, which property 
they formerly possessed. Again, the fungi obtained in one and the same 
