Biologic Forms of Erysiphe in Cases of Non-Infectioii22 1 
that the cases of non-infection described above are due to the 
inability of the haustorium to carry on its normal functions under 
the conditions obtaining in the cell, i.e., they are evidences of defective 
symbiotic arrangements between the parasite and host. 
An interesting variation in the behaviour of the germinating 
conidium of E. Graminis was observed. As a rule, when the 
conidium germinates, the germ-tube at once pierces the epidermis, 
and forms a haustorium in the epidermal cell. The germ-tube can 
often be seen passing over an open stoma, as in Fig. 2, where the 
haustorium has been sent into the subsidiary cell. No case has 
hitherto been recorded of the entry of the germ tube of the conidium 
through a stoma, all authors having described only the process of 
direct penetration of the cuticle. In the present experiments, a 
very few cases 1 were observed in which the germ-tube of the 
conidium, instead of passing over the open stoma, entered it, and 
growing inwards through the respiratory cavity, had come into con¬ 
tact with the mesophyll-cells (Fig. 1). At this stage the experiment 
had been concluded by the fixing of the material, so the further 
growth, if any, of the germ-tube could not be observed. Whether 
there is any infection in such cases, 2 3 4 5 6 7 8 and whether conidia of 
E. Graminis show similar behaviour when sown on their “ right ” 
host-plant, are points which must remain doubtful for the present. 
This entry, in exceptional cases, of the germ-tube through the 
stoma is of interest in connection with the fact, which I have lately 
recorded (5), that under cultural conditions, E. graminis possesses 
the power of growing as an endophytic parasite. 
1 Six cases, in all, were observed among the many hundreds of 
germinating conidia seen. 
2 In the case of uredospores, Miss Gibson has shown, in a recent 
paper in this journal (Vol. III., p. 184), that entrance of the 
stoma by any germ-tube is no index of the capacity of that 
germ-tube to infect the leaf. 
Bibliography. 
(1) Marchal, E. ; in Comptes Rendus, CXXXV., 210 (1902), and 
CXXXVI., 1280 (1903). 
(2) Salmon, E. S. ; in Beihefte z. Botan. Centralbl., XIV., 261 
(1903), and in Journ of Bot., XLI., 159, 204 (1903). 
(3) Idem; in Annal. mycolog., II., 255, 307 (1904). 
(4) Idem ; in Annals of Bot., XIX., 125 (1905). 
(5) Idem ; in Phil. Trans., CXCVIII., 87 (1905). 
(6) Ward, H. Marshall; in Proc. Roy. Soc., XLV1I., 441 (1890). 
(7) Idem ; in Annals of Bot., XVI., 299 (1902). 
(8) Massee, G.; in Phil. Trans., CXCVII., 7 (1904). 
[Tor Description of Plate V. see over. 
