and of some allied Scrophu lariaccae. 395 
hairs on the more peripheral portions of their lower faces. 
Pedictdaris sylvatica grows in moist meadows. Its lowest, 
quite subterranean, leaves have backward ly recurved margins, 
and only under the reflected portions on the lower face are 
the numerous dome-shaped glands which are mingled with 
capitate hairs. Under these dome-shaped glands the fine 
nerves form an almost subepidermal network of tracheides. 
The foliage-leaves have recurved margins, and possess 
numerous dome-shaped glands on their lower faces, along 
the courses of the fine nerves : but the glands are not confined 
to the reflected portion. Tozzia alpina grows in wet alpine 
spots. The plane foliage-leaves possess on their lower faces, 
along the courses of the fine nerves, numerous dome-shaped 
glands and capitate hairs, which are especially abundant near 
the marginal sinuses. The wholly subterranean scales have 
their lateral margins very strongly reflected backwards, so 
that only small parts of their lower faces are visible. 
Numerous dome-shaped glands and capitate hairs clothe 
the inner (lower) surface of the reflected parts, and many 
dilated tracheides end directly under the epidermis of this 
region. The scales higher up have their lateral margins 
strongly reflected only on the more distal parts, and their 
basal portions are like the foliage-leaves as regards the dis- 
position of the hairs. Thus these three Scrophulariaceous 
plants (the only ones on which I can find observations 
recorded with reference to subterranean scales and foliage- 
leaves), growing in moist or wet soils, have a specially 
well-developed system of dome-shaped glands. So far as 
I can judge from Hovelacque’s descriptions and figures, it 
seems that there is a relative increase in the development of 
these glands in the subterranean scales compared with the 
aerial leaves : this is very clearly marked in Tozzia and in 
Lathraea. This recalls the increased development of the 
physiologically equivalent water-excreting mechanism to 
which I called attention in the subterranean parts of sapro- 
phytes (8). It further leads us to anticipate that ordinary 
green geophilous plants will possess some similar mechanism. 
