July 15, 1914 
Fusarium on Sweet Potato 
253 
these species. The spore colors were emphasized through the contrast 
with the carmine mycelium thallus common to all four strains. These 
ochreous and orange spots contained masses of characteristic sickle¬ 
shaped conidia. One culture was found to have ochreous spore masses 
of the type illustrated in Plate XVI, fig. /, although the majority of 
cultures contained spores of a longer type (Pi. XVI, fig. 0 ). Cultures 
with orange conidial beds, however, presented a quite different type from 
both of those just mentioned. This type, although not figured, was 
more subulate than the forms in Plate XVI, fig. M. A second orange 
type in another culture was of the same general shape, but a little broader. 
(See PI. XVI, fig. G.) The descendants of single spores of these four types 
of sickle-shaped conidia, repeatedly transferred to different vegetables, 
stems, heads, and grains, remained distinct from each other, but constant 
in themselves in the same spore stage. The sporodochia were formed 
more freely in subcultures, especially on stems. In addition to this 
so-called tubercularia-like sporodochial stage, one of the four strains 
! (PI. XVI, fig. 0 ) produced perithecia and was identified as Gibberella 
Saubinetii (Mont.) Sacc. It had no chlamydospores, but thick-walled 
swollen cells occurred in plectenchymatic bodies (stroma) and in closely 
interwoven hyphae (PI. XIV, fig. /), which often resembled chains of 
chlamydospores, but differed from them in function. The other similar 
type (Pi. XVI, fig. J) never formed perithecia, but produced clusters of 
chlamydospores. The latter were quite as resistant as perithecia and 
havej'prov^ to be an effective resting stage in species destitute of 
the ascigerous form. Later, the two other strains of the section 
Roseum with orange conidia were also found to differ more than was 
apparent at first glance. The broader type produced blue globose 
sclerotia in cultures on stems repeatedly watered when dried out. These 
bodies resembled in general appearance perithecia of Gibberella, but 
were massed and much smaller, 56 to 70/1 in diameter, and without any 
indication of asci. Such sclerotia were entirely wanting in the species 
having the slender type of conidia. The study for some years of these 
four above-described strains isolated from wheat ( Triticum spp.) did 
not change the previous conclusion that they were four separate fungi: 
Gibberella Saubinetii (Mont.) Sacc. (PI. XIV, and PI. XVI, fig. O); 
Fusarium culmorum (W. G. Sm.) Sacc. (PI. XVI, fig. J) ; F. subulatum 
App. and Wollenw.; and F. metachroum App. and Wollenw. These 
fungi were formerly included in the collective species F. roseum Link 
autorum. 
From this example it is seen that to differentiate these fungi is only a 
question of method. A skeptic, however, may ask, Why is Fusarium 
culmorum not the conidial stage of Gibberella? It must be conceded 
that the conidia derived from ascospores of Gibberella resemble in shape 
those of F. culmorum (PI. XVI, figs. J and 0 ). It would not be con¬ 
vincing to state that F. culmorum has 5-septate conidia with a maximum 
