34 BOTANICAL GAZETTE [JULY 
or in the sheltered places among the thick tops. Here it starts in 
scattered spots, each very distinct, and if uninterrupted continues 
spreading until late November, passing into the teleuto stage in the 
ordinary manner. The feature shown in figs. 14 and 15 is, however, 
one of the most remarkable. In this particular instance the rust 
started in the uredo form in the scattering manner just described. 
Fig. 14 is in a mass of tops sheltered by taller growth toward the west, 
Fic. 12.—Old slough bed at Sacramento, with as 
of rust. August 14, 1903. 
paragus to left; starting point 
f 
and jig. 15 is the east end of an east-and-west row. Just as this was 
well started (the condition all over the district was mostl 
the wind revived in a very dry form and blew steadily for a number: 
of days, with quite cold nights. Immediately the uredo rust on the 
green stalks turned to teleuto, the rust stopped spreading, and the} 
fields looked exactly as though a fire-brand had been thrust into the , 
green tops here and there, producing a black, dead spot in the green, 
healthy growth. The to 
up through, and the fiel 
dead, teleuto-covered patches, surrounded by a 
y the same)\, 
nd in contact with 
ps being still growing, new growth came’! 
ds were spotted with these perfectly black, \ 
\ 
