474 BOTANICAL GAZETTE [DECEMBER 
mesophytic vegetation (with Aira cespitosa, Ranunculus acris, etc.). No dif- 
ferences of moment were observed between east and west slopes. Although no 
close study of factors was made, it was clear that the direct influence of radiation, 
or its influence on moisture and temperature, is the controlling element. A figure 
is given of a valley in Lapland with a wood of birch with scattered pine and 
spruce on the southern slope, and a tundra with Salix herbacea, Andromeda 
hypnoides, and other characteristic forms on the northern slope—H. C. Cow es. 
SALMON has extended his investigations on the specialization of parasitism 
in the Erysiphaceae to include the ascospores of several forms.?7 In former 
experiments with conidia it was found that conidia from Bromus commutatus 
would infect B. hordeaceus but not B. racemosus or B. mollis. B. hordeaceus, 
however, acts as a ‘“‘bridging species,’’ so that if conidia from B. commutatus are 
used to inoculate B. hordeaceus, the conidia produced on B. hordeaceus will in 
turn infect B. mollis. <A similar specialization occurs in the ascospores. Asco- 
spores from B. commutatus infect B. hordeaceus, but not B. racemosus. The 
conidia produced from the former infection were then used to infect B. mollis. 
nother series of experiments was performed in order to see whether E. 
graminis from wheat, which also infects Hordeum silvaticum, could by continuous 
culture on H. silvaticum be made to lose its power of infecting wheat. S 
fungus was cultivated for five generations on Hordeum, but showed no signs of 
losing its power of infecting wheat. Successive generations of the fungus on 
Hordeum seemed to show a weaker power of infecting that plant, but this is 
probably explained by the fact that only the younger leaves of Hordeum are 
susceptible to infection —H. HAssELBRING. 
and TANSLEY give a detailed account of their methods for surveying 
vegetation, applicable where the ground is reasonably flat.22 The two methods 
employed, the method of squares and the gridiron method, are related, and their 
especial advantage is to exhibit the characteristics of an area where the vegetation 
is complex and yet definitely related to physiographic features. These methods 
have been employed in the salt marshes of the north coast of Brittany, and in 
some respects resemble the quadrat methods used by CLEMENTS. The method 
of squares is employed for mapping on the scale of 1:250 or 1:500, squares one 
hundred feet each way being made from a base line; such squares correspond to 
a five-inch square in ea —— The gridiron method is used for detailed 
work, an area twenty-fi being out; this area is then divided 
off by tapes into smaller atia: See feet. Here six inches correspond to one- 
tenth of an inch in the notebook, so that great accuracy can be secured. The 
methods could be advantageously modified by the use of metric units and relations. 
27 SALmon, E. S., On ee of parasitism in the Erysiphaceae. III. Ann. _ 
Mycol. 3:172-184. 1905 
28 OLIVER, F. W., and Tans.ey, A. G., Methods of surveying vegetation on a 
large scale. New Phytol. 3: 228-237. 1904 
