1904] BRIEFER ARTICLES 463 
dead (protective ?) hairs and knob-shaped partly sunken glands; the 
inland form showed only the glands. None of the true halophytes 
showed slender dead hairs, though a few, such as Luda marina, possess 
glandular hairs. Artemisia Stellertana, found in the same habitat as 
Convolvulus, has a dense covering of long silky hairs. 
. The stomata were examined with respect to distribution on the 
two surfaces, level with regard to the surface, and form. ‘The only 
significant point observed is the presence of a few stomata on the upper 
surface of Convolvulus in the maritime form, while stomata are confined 
to the lower surface in the inland form. ‘This observation does not 
support the accepted view that conditions on the maritime strand are 
essentially xerophytic, for so-called xerophytic leaves usually have 
stomata only on the under surface. Moreover, out of twenty-four 
species found near the shore at Woods Hole, all but three (viz., Ligus- 
ticum scoticum, Artemisia Stelleriana, and Sabébatia steliaris) have stomata 
on both surfaces. In seventeen species the stomata are level with the 
surface, in three sunken less than half the thickness of the epidermis, 
in four (viz., Euphorbia polygonifolia, Atriplex hastata, A. arenarta, and 
Polygonum maritimum) sunken half or more than half the thickness of 
the epidermis. ; 
The differences noted in the foregoing paragraphs point with few 
exceptions to a more xerophytic structure in the leaves of the maritime 
specimens than in those of the inland specimens of the same species. 
Since most of the conditions on the strand of Lake Michigan are very 
similar to those on the Massachusetts coast, it is natural to look to 
presence of salt in the soil or in the air as the cause of the differences 
observed. To further test this point specimens of Sodidago sempervirens 
were collected from stations (1) as near as possible to the strand, (2) 
one hundred feet back from the water’s edge, on the slope of a hill, (3) 
on the edge of a marsh bordering a brackish pond. The leaves pre- 
sented the same structure in all three cases, but the average thickness 
was as follows: from (1) 0.50, from (z) 0.36", from (3) 0.56™". 
There are evidently several factors of which account must be taken in 
dealing with figures such as these, but the observation seems plainly to 
support the assumption that the amount of salt present influences the 
thickness of a leaf. Measurements were also made of leaves of Statice 
Limonium collected from four situations whose salt content was estima- 
ted by titration with silver nitrate. The results of this set of experi- 
ments show a general correspondence between thickness of leaf and 
saltiness of substratum, but the results are not altogether trustworthy 
