ON THE MORPHOLOGY OF HEMILEIA VASTATRIX. 5 



After growing thus for some time with a sinuous course and 

 uniform diameter the germinal tube dilates, at some place, usually- 

 near the growing end, into an ovoid or pyriform sac-like vesicle, 

 into which all, or nearly all, the coloured contents soon pass 

 (figs. 13 and 13), leaving the rest of the tube and spore empty 

 of everything, except a few granules and frothy vacuoles. 

 Pig. 11 y shows in outline what an extensive growth may take 

 place before this sacculation occurs ; it usually occurs sooner, 

 however. 



The pyriform dilation may remain simple, or put forth branch- 

 ing processes here and there from any point (fig. 14). Sometimes 

 it grows forwards as a simple tube (fig, 15), on which a similar 

 dilatation may arise afterwards, and in this case the coloured 

 contents pass forwards into the new vesicle. This forward 

 growth is very rapid, and accompanied by vigorous streaming of 

 the protoplasmic contents. It sometimes happens that a septum 

 is formed across the neck of this swelling between it and the 

 rest of the tube (fig. 11). 



Where the branching is vigorous these vesicular bodies may 

 become very complicated, and assume the most grotesque 

 figures ; huge vacuoles, streaming, &c., arise as the growth con- 

 tinues (fig. 11). This is not for long, however, and though 

 the swellings may remain some hours after the remainder of the 

 tube and spore have rotted, they and their contents become at 

 length the prey of Bacteria, Torulse, &c. 



On sowing the Uredospores on the lower surface of vigorous 

 living coffee leaves, I obtained results in the main similar. The 

 germinal tubes produced, however, are usually shorter and less 

 branched, or quite simple, and the whole process is apparently 

 carried on more energetically. 



At fig. 16 is represented a piece of the lower epidermis of a 

 cotyledon of Cojfea Amhica, on which Uredospores had been 

 sown some eighty hours ; the spores germinated and put forth 

 the tubes freely as described above. The pyriform vesicle 

 appears very early, and receives the whole of the contents. 

 And now the meaning of the vesicular swelling above described 

 becomes clearer, for it is usually formed over the orifice of a 

 stoma and sends its processes through this into the intercellular 

 spaces of the leaf (PI. II, fig. 18). This is, in fact, the act of 

 " infection." The IJredospore on germinating produces a simple 

 mycelium (the germinal tube), which grows rapidly at the ex- 

 pense of the reserve material in the spore, and is only capable 

 of further progress on reaching the interior of the leaf in this 

 manner. 



The commencement of this further development is well shown 

 in figs. 19 and ^0, That the internal mycelium is simply an 



