and of some allied Scrophu la riaceae. 391 
sods, and cut shoots into which water was forced under 
pressure, when kept in a saturated atmosphere, both gave 
the same results as regards the excretion of water. Water 
was slowly poured out by the young leaves. The observa- 
tions on the source of the excreted water were at first quite 
unexpected. Rarely could any water be seen on the lower 
faces of the leaves (where the dome-shaped glands occur) ; 
whereas it was constantly visible in the sunken channels 
marking the upper faces of the secondary nerves, and extended 
as a thin film to the marginal sinuses. At first it appeared 
that this water could not have been excreted by the dome- 
shaped glands: nor did it seem more likely that it could 
have been poured out by the capitate hairs or stomata, both 
of which are more numerous on the lower than on the upper 
face of the leaf. A few simple experiments cleared up the 
mystery. A shoot was held in an erect position with the 
leaves extended in the normal position, and a drop of water 
was placed on the lower face of an unmoistened leaf near the 
margin. The drop spread to the nearest marginal sinus, and 
extended itself in the form of a thin film along the sunken 
channel marking the upper face of the secondary nerve which 
supplied that sinus. A drop of water placed on the lower 
face of a leaf near the midrib (where no glands occur) did not 
move : it remained a drop. Finally, a drop of water placed 
on the upper face of a normally disposed leaf flowed along 
the sunken channels of the secondary nerves towards the 
channel of the midrib, but did not appreciably spread towards 
the margin. These experiments show, therefore, that in a 
normally excreting leaf, the water present in the peripheral 
portions of the nerve-channels in the upper face must have 
been excreted by the upper or lower face near the margin. 
Inasmuch as the histological details of the upper face of the 
leaf are essentially the same throughout, and there is no close 
connexion between the upper epidermis and the tracheides, 
nor any peculiarity in the epidermis near the margin, there 
is every reason to believe that the water is not excreted 
by the upper face. On the contrary, the portions of the lower 
