38 PHYLOGENY 



and numerous fruit trees owe their large number of varieties, 

 and the high degree of constancy of all the samples, even, to this 

 exclusion of one of the two main parts of the common variability. 

 The term variety is often taken as indicating only one single indi- 

 vidual with its peculiar characters acquired in the embryonic stage, 

 which remain unchanged in the thousands and in the millions of its 

 grafts and cuttings. 



Nature, of course, makes some use of vegetative multiplication, too. 

 But since this process does not play any prominent part in the cur- 

 rent theories concerning the larger features of progressive develop- 

 ment, it has no further interest for our present discussion. 



One main point, however, has to be considered. It is seen by 

 looking at the question from the opposite side. In order to take a 

 definite example, we may ask to what extent an observed character 

 is due to embryonic variability, and which part falls to the partial 

 variability. In all cases of branched plants there is no difficulty, and 

 every one will grant that only the average of all the leaves or all the 

 fruits or all the flowers can be the result of embryonic changes. 



But in the case of main stems and main roots there is no possibility 

 to determine this average. An annual plant has one main stem, 

 whilst a perennial species has many of them. The one stem is 

 obviously to be considered as equivalent to only one of the many in 

 the latter case. It may be of average height, but it may as well be a 

 more or less extreme variant. 



Exactly so it is in practice. The amount of sugar in an individual 

 sugar-beet is partly due to embryonic variability and partly to the 

 subsequent influence of treatment and weather. Now it is manifest 

 that both are of value for the direct industrial purposes, but it is 

 equally manifest that both cannot have the same signification for 

 the value of the seeds which this individual plant may afterwards 

 produce. Or, in other words, the sugar-amount of a beet is in no way 

 a full and reliable indication of its value as a seed-bearer. If a high 

 amount of sugar is due to embryonic variability, it is indicative of 

 high excellence, and will probably be followed by the production of 

 seed of primary quality. If, on the other hand, the percentage figure 

 is reached by exceedingly favorable weather, or by an accidentally 

 good position of a plant on the field, as to light, space, and the escape 

 from all its enemies, it is no indication at all of embryonic variability, 

 and it is very dubious whether it is to have a lasting influence on the 

 seed, in fact, such a relation is strongly denied by "Rimpau, von 

 Ruemker, and other German authorities, whilst it was believed in by 

 the famous English wheat-breeder Hallett. 



Granting the exactness of the first view, the sugar-percentage 

 figures are seen to be reliable only when subsequent or partial 

 variability is sufficiently excluded. A hundred of selected beets may 



