402 Scott & Scirgant. — On the Development of 
a smooth white epidermis. It must be remembered that the 
insertion of the cotyledon divides the surface of the tuber 
into two regions. The epidermis covering the lower part 
is continuous with the outer epidermis of the cotyledon, and 
ceases abruptly where the primary root begins. The upper 
part of the tuber is enclosed within the cylindrical cotyledon, 
and the epidermis which covers it is continuous with that 
of the inner cotyledonary surface and with the epidermis 
of the first leaf. The formation of periderm begins in 
the layer of cells immediately below the epidermis of both 
regions. Transverse sections through such a tuber as that 
drawn in Fig. 4 show five or six layers of periderm-cells 
near the middle of the tuber, and two or three on either side 
of the zone in which the cotyledon is inserted. These two 
formations are connected by one or two layers of periderm- 
cells which cut right across the parenchymatous tissues of 
the cotyledon, just where it is inserted on the tuber. In older 
seedlings no doubt a greater number of cell-layers is formed. 
The two parts often remain in contact with one another for 
some time, though organically separated. It is often possible 
to dig them up without shifting them. In the summer 
new roots are sent out (from four to six in number) from 
the upper surface of the tuber (Fig. 6). Some of these, 
generally two, are contractile 1 (see Fig. 8). We found root- 
hairs produced in clusters near the tips of the contractile 
roots at this stage, by which, no doubt, they are firmly fixed 
in the ground. It is clear that without some point of attach- 
ment the effect of the contraction would be to draw the tip 
of the roots up, rather than to pull down the tuber. But 
measurements show that the tuber actually sinks deeper 
into the ground, and in this way is at last freed from its 
discarded cotyledon and seed-coats, which are left to rot 
nearer the surface. 
The result of the contraction of the rootlets is very startling. 
The tubers when last examined in May (Fig. 4) were only 
1 Rimbach, Berichte der Deutsch. Bot. Gesell., April, 1897 ; and Prof. F. W. 
Oliver, Journal of the Royal Horticultural Soc., April, 1898, p. 493. 
