HEPOET ON VAUIOrS MANrEES AT CHISWICK. 
53 
The noticeable points in Achillea were the proportionally large 
number of runners in no. 1, combined with the absence of stock- 
branches, and the frequent disappearance of the primary root. 
In box 2 the stock-branches were somewhat more numerous, but 
there was a great paucity of runners ; the main roots were pro- 
portionally well developed. In 3, runners were numerous and 
long, stock-branches scanty, main roots large. 
Owing to the variety of siructures to be taken into account, 
a general comparison of roots and plant is more than ordinarily 
*• difficult to be obtained. The following is an approximate esti- 
mate of the condition of the plants in October, and the roots 
proper in April. The most noteworthy circumstance is the an- 
tagonism between the roots and the plants in 1, 2, and 3, the 
plant in the unmanured box exceeding those in the manured 
boxes : — 
1 
2 
3 
4 
5 
6 
a 
h 
c 
/ 
e 
d 
Plant 
f 
e 
d 
e 
I 
a 
xii. Carum Carui. 
Box 1. Hoots pretty uniform. Eoot tap-shaped, vertical, 12- 
14 inches long, | inch in diameter, at the upper end tapering olf 
very gradually towards the apex, yellowish-brown, marked above 
with circular ring-like prominences, and throughout its whole 
length by small scattered elevations, from which the branches 
proceed smooth below; branches 1-2 inches long, given off 
obliquely downwards at tolerably regular intervals of j to 1 inch 
above, less numerous and more widely distant from each other 
below. The pseudo-nodes are arranged in four vertical lines, so 
that the fifth is over the first. The fibrils are few in number, 
proceeding from similar raised pseudo-nodes as the larger branches, 
sometimes in company with them, at other times distinct, the 
upper ones ascending, the lower ones obliquely descending. 
Box 2. Eoots varying greatly in size, some penetrating the 
bottom of the box; the largest 18 inches long and | inch in 
diameter at the thickest portion; average roots scarcely, if at 
all, larger than in no. 1. The fibrils more numerous than in 
no. 1. 
Box 3. Variable, but less so than in no. 2; length 14-1& 
inches, J-| inch in diameter at the widest part, Some of the 
