32 E. J. BUTLER. 
pitted (the appearance is shown in an exaggerated form in plate I, 
fiz. 4), but I have not been able to satisfy myself as to the exact struc- 
ture of the markings seen, and they are not visible in optical section 
which tells against their being pits. Immature spores are colourless 
and with the outer wall little developed. The spores measure 
10 to 15°5 by 10 to 15,, usually about 12,, if globose. The average 
of 100 spores from an agar culture of the cotton fungus gave 13°5 by 
12,. Fifty-four spores measured directly from pigeon-pea gave 
10 to 14:5 by 10 to 12°5,, the average being 11°9 by 11°5,. Smith 
gives 10 to 12,, if globose, and 8 to 12 by 11 to 14,, if ellipsoidal, and 
notes that spores from a potato culture of the cow-pea fungus were 
generally larger than 12, in diameter, and also larger than those 
measured directly from the host plant. In all the ascospore 
characters there is the closest possible agreement between the 
Indian and American forms, if we except the slight markings of the 
spore wall, which are not mentioned by Smith. They are only 
visible under high powers and are sometimes very difficult to 
make out, so that much importance does not attach to them. 
The conidial stages are, as pomted out above, practically value- 
less for purposes of comparison. Their variability is such that it 
is hardly possible to separate by microscopic examination and 
measurements alone many of the different species of Fusarium from 
one another, and still less those of Cephalosporium. Smith gives 
the spores of the former stage of Neocosmospora vasinfecta as 
ijunulate, 3 to 5-septate, and 30 to 50 by 4 to 6u in diameter, and of 
the latter as oval to narrowly elliptical, straight or slightly curved 
and 4 to 25 by 2 to 6. I have found that the majority of the 
Fusarium spores are from 35 to 40, long, but the lower limit is 
very elastic, as all gradations down to the Cephalosporium type 
can be found. The breadth of the spores and the number of septa 
agree with the data given by Smith, but the intermediate types 
between Fusarium and Cephalosporium often have one or two septa. 
These types develop chiefly in culture, the spores formed in 
stromata on the bark of the host being more regular. The 
Cephalosporium spores average about 6 to 10 by 2 to 4y, and 
