300 
W. B. Mercer, 
complex forms are sometimes suggestive of chains of brown coniclia 
(Fig. 4, /, 5 ). 
b) The second variety is represented by specialized stretches of the 
mycelium (Fig. 4, 9 — 16 ). It has already been mentioned that in old 
mycelium the protoplasm becomes concentrated locally in bands of 2—10 
cells’ length. The cells divide, swell, thicken their walls, and assume a 
dark colour. Short branches with unthickened walls often arise from them 
(Fig. 4, 16 ). 
c) Gemmae may also take the form of grape4ike masses of cells 
(Fig. 4, iy — 21 ). They may be terminal (Fig. 4, iy — 19 ) or in the middle 
of a hypha (Fig. 4, 20 , 21 ). Their development was followed in the usual 
way, and takes place in the repeated division of a few adjoining cells, 
with concurrent swelling of the separate divisions. The walls gradually 
thicken and darken in colour; for the most part the masses have reached 
a considerable size before the thickening and darkening becomes apparent 
(Fig. 4, 22 a — e). This form of gemma is very characteristic of Salep 
agar cultures, and is comparatively rare on plum agar. 
Combinations of the different forms of gemmae occur, e. g. a conidium- 
like growth arising from a gemma of type b or c (Fig. 4, 12 , /j); there 
are many intermediate forms between types b and c (Fig. 4, 16 , 2 j); 
while some are so irregular (Fig. 4, 6 , y, 8 ) that it would be difficult to 
place them in any of the three classes. 
All forms agree in being dark-coloured, the colour deepening with 
age, and in many cases becoming finally black. In old cultures gemmae 
are often impenetrable by light, even after boiling in Lactophenol, so that 
the structure cannot always be seen. They contain variable quantities of 
oil. With prolonged soaking in Lactophenol the colour is partially removed 
and they become transparent, many of the oil drops becoming at the 
same time invisible. In gemmae of all types it is frequently to be 
observed that the individual cells round themselves off, and in so doing 
partially separate (Fig. 4, 5 — 7 , 12 ). 
A halo of dark-coloured agar usually surrounds these bodies. In a 
moist-chamber culture it may be observed gradually increasing in density 
with the growth of the gemma; it is always darkest nearest the gemma, 
becoming gradually lighter in shade towards the periphery. On Salep 
agar the colouration is not so marked as on plum agar, and on gelatine 
media it is very faint or altogether absent. 
The gemmae are all capable of germination. When the agar of 
an old culture is teased out and spread over fresh agar they germinate 
readily. Germination takes place in the pushing out of one or more 
germ tubes, whose bases are frequently swollen (Fig. 4, 24 — 2 y). Usually 
the number of germ tubes produced is small. Germination also takes 
place, though more slowly, in hanging drops of water. 
The simpler forms of type a are so similar to the brown conidia 
on the aerial mycelium, that they might well be regarded as conidia, 
distorted by reason of the fact that they are produced in the agar. But 
the gradation of forms from the simple to complex irregular types with 
no morphological resemblance to conidia is sufficiently complete to justify 
their inclusion together as gemmae. 
The development of the third type follows the same lines as that 
oft he meristogen pycnidia, and the gemmae are often fairly regular; 
