102 VARIABILITY 



earliest variations were spontaneous, for it is at least unlikely that 

 the multiplication of the primitive living beings involved exact 

 quantitative and qualitative division. Doubtless early variations, 

 however caused, were seized upon by Natural Selection, which, as 

 a necessary antecedent to all else, established the insusceptibility 

 of the germ-plasm and regulated the tendency to vary in all 

 directions about the specific mean. In other words, though 

 Natural Selection cannot have been the original source of varia- 

 tions, it has established variability as an adaptation and ensures 

 its constant occurrence, and in each species has regulated the 

 amount of it according to the needs of the time. At any rate, 

 however spontaneous variability originally arose, it certainly 

 exists and presents all the signs of an adaptation controlled by 

 Natural Selection. 



165. Striking evidence that spontaneous variability is a 

 strictly regulated adaptation, and that germinal insusceptibility to 

 the direct action of the environment is also an adaptation, is 

 furnished by four well-authenticated facts. First, plants propagated 

 asexually (by slips, suckers, etc.) vary very little as long as they 

 are so propagated, no matter how numerous the pseudo-generations, 

 nor how diverse the influences to which they are exposed and by 

 which they may be modified ; wherefore gardeners, whenever 

 possible, propagate valuable varieties by means of cuttings. 1 This 

 indicates that variations are, at least, rare amongst somatic cells, 

 for, if they were common, plants propagated asexually would be 

 changeful. Second, offspring arising by seminal generations (i.e. 

 from germ-cells) always vary from their parents. This indicates 

 that variations occur normally within the limits of the germ-tract, 

 that is in those lines of cells by which the germ-cells of the 

 individual descend from the fertilized ovum whence he sprang. 

 Third, the offspring of a plant reared from its seed vary no more 

 and no less, apparently, when the seed is gathered at the end of 

 the first season, than when it is gathered after many seasons during 

 the course of which the plant has been propagated asexually. In 

 other words, the lengthening of the germ-tract by any number of 

 pseudo-generations (i.e. slips and suckers) does not appear to 

 increase the number and magnitude of variations. This indicates 

 that variations do not arise all along the germ-tract, but once for 

 all in some particular part of it ; for, if they arose all along the 

 tract, the lengthening of it would tend to influence their number 

 and magnitude. Fourth, ' identical ' twins, which originate from a 



1 See 183. 



