342 GEORGE MASSEE [NOVEMBER 
stem before the bulb begins to form. This infection I consider to be 
due to the mycelium present in the testa of the seed. Unfortunately 
complete proof of this is not forthcoming as it was in the tomato 
disease described above, as up to the present moment no one has 
succeeded in causing any member of the Uredineae or Ustilagineae to 
produce fruit as a pure culture, or apart from the natural substratum. 
In many instances the mycelium passes up the flower-stalk and 
enters the anthers and ovary without however producing spores, this 
final act being prevented by conditions at present unexplained. 
Mycelium can be frequently observed in the testa of seeds produced 
by such plants, and if the seeds are sown under conditions preventing 
external inoculation many diseased plants result. 
Plants attacked in this manner can be easily recognised after a 
little experience, owing to the deep blue-green colour of the flower- 
stalk. 
Returning to the experiment with wheat. Distinctly shrivelled 
grain caused by the presence of the rust called Puccinia glumarum, 
Eriks. and Henn., better known in this country as Puccinia rubigo-vera, 
DC., developed on the chaff, and sometimes also on the grain itself, 
was used. Forty grains were sown in each of two pots, one containing 
ordinary soil, the other rich stable manure with a small admixture of 
soil. Each pot was protected by a glass vessel with cotton-wool, as in 
the experiments described above. In the pot containing ordinary soil 
sixty per cent of the grain germinated, whereas in the richly manured 
pot only fifty-two per cent germinated. When the plants were three 
inches high indications of rust pustules were seen on a few leaves in 
each pot, and when the plants were five inches high twenty-six per cent 
of the plants were rusted in the pot containing ordinary soil, and forty- 
seven per cent in the richly manured pot. At this stage the experiments 
terminated, as the spores were in some instances mature, and the 
plants being crowded inoculation from spores would probably have 
taken place, and thus a greater percentage of rusted plants would have 
resulted than those due to what I consider as cases of rusting from 
the use of diseased seed. As a control experiment a similar number 
of plump and healthy grains obtained from plants having the foliage 
badly rusted, but the ears perfectly free from rust, were sown in 
ordinary soil and protected as described above. Ninety-six per cent 
germinated, and all the plants remained perfectly free from rust. 
Another experiment was conducted as follows:—A jar was filled 
with sterilised water containing a small amount of extract of manure ; 
a piece of coarse muslin was stretched over the top of the jar just in 
contact with the quid. Twenty grains of wheat obtained from a 
plant not attacked by rust and presumably healthy were placed on the 
muslin, the jar being protected by a glass globe. Nine of the grains 
produced vigorous plants, the remainder being weakly were removed. 
When the plants were about two inches high a sterilised piece of 
