6 H. MARSHALL WARD. 



extension of the germinal tube is proved by sucli examples as 

 fig. 31 : the spore germinating close to the orifice of a stoma, 

 has sent its tube directly through into the leaf without forming 

 the preliminary vesicle. These and many similar preparations 

 were obtained by sowing spores on living leaves kept damp in 

 glass cells, and cutting vertical sections at periods varying 

 from 24 to 100 hours afterwards. 



I carried still further the proof of the fact that the internal 

 mycelium is but an extension in the leaf of the germinal tube 

 by sowing spores on the tipper surface of leaves at places 

 from which the epidermis had been removed ; the result 

 was a rapid growth of the germinal tube directly into the 

 tissues, pushing its way between the palisade cells as it advanced 

 (figs. 22 and 23). Here again no vesicle was formed. All 

 attempts to infect by sowing on the uninjured upper surface 

 have failed ; the spores germinate, tubes and vesicular swellings 

 form as on glass, but the whole soon shrivels and dies. 



The mycelium within the leaf, then, the action and extension 

 of which corresponds to the yellow discoloration seen externally, 

 is clearly but a continuation of the germinal tube sent forth 

 from the Uredospore, and which enters a stoma as described. 

 Once established in the lacunae of the leaf this soon branches, 

 chiefly at first in the plane of the leaf, and feeding upon the 

 products of the cells of its host, produces the injury.^ At first 

 the young mycelial tubes are very delicate, filled with fine- 

 grained protoplasm, and somewhat stumpy; they soon become 

 vacuolated, and more coarsely granular, and send out tufts of 

 short, thick branches towards the cells bounding the intercellular 

 spaces, while here and there longer *' leaders " run out 

 between the cells in various directions. 



The main features of the internal mycelium thus produced 

 are typically as follows. Its ramifications are confined to the 

 intercellular spaces (fig. 24), except that at numerous points here 

 and there very slender processes pierce the cell-walls to form 

 haustoria. The mode of branching is extremely irregular, and 

 influenced by the arrangement of the cells between which the 

 branches run ; the rate of growth, depending on several circum- 

 stances, also affects the length of the branches. 



Transverse septa occur here and there, often separated by long 

 intervals (fig. 28), especially at the peripheral parts. The 

 diameter of a mycelial thread is about the same as that of the 



* As further evidence, I may remark the success of infection experi- 

 ments based on these observations. In one instance, I made sixteen 

 separate sowings on healthy leaves of as many plants : in fifteen cases the 

 " disease spot " appeared where the soviin^ was made, and nowhere else on 

 the plant. 



