THE WILT DISEASE OF PIGEON-PEA. 47 
clustered conidiophore of the microconidial stage, which, however, 
is not a constant character in culture. The ‘species named Fusa- 
rium Dianthi Pril. and Del.(17), the cause of the carnation disease 
in the south of France, is evidently closely similar to the conidial 
stages of Neocosmospora, since Delacroix in a subsequent paper(5) 
stated that they appeared to be identical. This same fungus was 
referred by Mangin(18) to Fusarium roseum, one of the commonest 
members of the genus, and hence this must also be considered as a 
close ally of the pigeon-pea parasite. Similarly, Fusarium oxy- 
sporum Schlect., with its numerous possible synonyms, is hardly to 
be distinguished morphologically from this species except in the 
more regular, coloured chlamydospores. The latter are, however, 
very variable bodies. Fusarium Lini Bolley (19), another of the 
proved wilt-producers, isalso probably a close ally. It is true that 
Bolley does not specifically mention a microconidial stage, but 
from his figures on pages 37 and 45 it would appear probable that 
one exists. 
In the absence of the perfect fruit it appears therefore necessary 
to rely on cultural characters for some of the chief diagnostic features 
of the pigeon-pea parasite. 
Cultural characters of the pigeon-pea parasite. 
The following media were inoculated from an agar tube culture 
originating from a single spore and being part of the material used 
for the successful inoculations in Series XIV. 
Nos. | to 10 consisted of a standard solution made up of ammo- 
nium nitrate 10 grms., potassium phosphate 5 grms., magnesium 
sulphate 2°5 grms., and distilled water 1,000 c.cs., to which were add- 
ed the substances specified below. The liquid cultures were in Pas- 
teur flasks containing 30 c.cs. of the medium, the solid cultures in 
tubes. 
A. Liquid media. 
1. Glucose 3% added to the standard solution. 
2. Saccharose 37 i Pe F 
