398 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
than portions of the plant which died down and apparently 
ceased to exist in the previous autumn. Hence the life of a 
single Potato plant may be prolonged year after year until 
through weakness or deterioration it comes to an end. It will 
therefore be obvious that improvement by selection of the tubers 
is impracticable. Anyone can demonstrate this by planting a 
tuber which from any cause may be misshapen. The produce 
No. 5.- Potato Beiuues and Foliage. 
Fig. 34. 
will revert to the uniform type of the variety to which it belongs. 
The only modification of this rule I am acquainted with is in the 
cases where all the tubei's of one plant show a unifoi'm diver- 
,i(('nc(! in character, either for better or worse. When this is so 
it is possible! that, by the selection of all or any of these tubers, 
a sHghtly different Potato might result, as in the case of some 
types of the Ashlcaf s(!ction, which are dwarfer ;ind more com- 
