ON VAaiETIES, lUCES, SUB-SPECIES, AND SPECIES. 71 



example of tlie first we have polatoes and the scales of tlie white 

 lily {Lilium candidum), and of the second the bulblets of Z/^7^«m 

 hulhiferum : these tubercular buds, if placed in earth, will produce 

 individuals identical with their respective parent plants. 



There are other methods of propagation by simple division, 

 but it is useless to mention them. 



B. Propagation by Seeds. 



Propagation by seed is capable of producing individuals very 

 different from those obtained by simple division. 



Indeed, although we generally find the greatest possible re- 

 semblance between a parent plant and one obtained from it by 

 any of t!ie above-mentioned processes of division, this is by no 

 means the case when we come to compare plants with their 

 seedlings : for although the latter have the characters that are 

 essential to their species, it may, we will not say always does, 

 happen that some of tlie individuals differ more or less from the 

 others. 



The consequences of this may be easily seen. Whenever it is 

 desired to propagate any one of the iimumerable plants in cul- 

 tivation, for the purpose of perpetual ing any particular quality, 

 recourse must be had to simple division; whilst, on the other 

 hand, if it be desired to obtain new varieties, that is, individuals 

 differing in some degree from their parents, recourse must be had 

 to sowing the seeds of the latter. 



A very good illustration of the foregoing remarks is afforded 

 by Rohinia pseud - acacia, var. spectabilis. M. Descemet 

 having sown in the early part of the present century some seeds 

 of Rohinia j)seud-acacia, obtained some seedlings without any 

 spines ;* and it is by the simple division of these seedlings that 

 all the specimens of the variety spectabilis have been obtained. 

 A very remarkable fact connected with this subject is, that the 

 seeds of the variety spectabilis have always produced spiny 

 plants identical with the common Robinia pseud-acacia. 



When we mention this fact we do not intend to say generally 

 that all seedling varieties yield seeds, which, when sown, pro- 

 duce plants identical with those from which the seedlings were 

 obtained ; but we think we may say that whenever we loish to 

 reproduce xoilh certainty any given peculiarity in a plant, we 

 must have recourse to simple division; for we cannot be cer- 

 tain to obtain ivhat we icaiit from the seeds of that plant. 



* I have called the thorns of this plant spines in accordance with the 

 usage of gardeners and many botanists, especially Desfontaines and Poiret; 

 but they are in fact prichles, being independent of the wood, and adhering 

 only to the bark. 



