FiuiguM of Lolium. 271 



section, for such cells are the only menil^ers of all the nucellar 

 tissue, which will stain in the same way as the funeral svsteni. 

 (Plat^XXI. Fig. 1). They stand out in contrast to the l)ackgrounii 

 of normal, unattacked nucellar cells. 



It is difficult to determine whether the hyphae actually 

 apply themselves to the nuclei, but it is readily seen that 

 the nuclei do undergo a definite change, l)ecomin<; large and 

 somewhat distorted, and at this stage will stain uniforndv 

 with the violet dye. These fungal-containing cells may occur in 

 any position in the nucellus, but at this stage they are few in 

 numl^er, and are generally to be found at the end of the ovule 

 furthest from the micropyle. They become more abundant after 

 fertilisation occurring in any part of the nucellar tissue. 



The embryo-sac at this stage is ready for fertilisation, and it 

 agrees with the rest of the ovule in containing the fungus. The 

 protoplasmic lining of the sac carries the hyphae. They run 

 rather sparingly along the sides, and at the distal end of the 

 embryo-sac, but are more abundant at the proximal end in the 

 vicinity of the ovum. Ibey are in close connection with both the 

 synergidae and the egg-cHl, and enter into the substance of, the 

 latter at this early point in the development of the grain. (Plate 

 XXI.. Fig. 6.) Previously it has been thought that " iuftction " 

 took place at a much later stage, after the differentiation of the 

 growing cone, when the formation of the embryo was fairly ad- 

 vanced. It has been suggested that the entrance of the fungus into 

 the embryo was due to the chemotactic influence of the growing 

 apex. My observations show that the fungal constituent is present 

 in the ovum before any divisions have taken place, and that the 

 formation of a special layer in the grain for the purpose of infect- 

 ing the embryo at any specified* period is not necessary. 



Stage B. 



Text- figure 6 illustrates the external appearance of the ovary 

 after fertilisation, and at the commencement of en<losperm forma- 

 tion. Ihe elongation of the ovary which accompanies this change 

 is beginning to be apparent. As in Stage A. it is drawn from the 

 ventral surface, and shows the same features as licfore. 



If an ovary be sectioned at approximately this stage, our know- 

 ledge of tlie relation of the fungus to the grass is considerably 

 augmented. 



The hyphae are still active in the carpel wall. The cells com- 

 posing the distal ju-ea of this wall are attacked by the fundus, their 



