36 



VARIATION. 



will be lowest, and all the other plants will be of intermediate 

 height, varying from the tallest gradually down along the row 

 to the shortest plant. If we draw a line through the tips of 



^s^- 



1q^ ^ »f # « 



Fig. 3. 



Above: a row of young plants of equal age, subject to greater or les- 

 ser influence of a scource of heat at the left. Continuous variation in 

 size. 



f' Below : Similar row with glass-plate separating plants at the left from 

 those at the right. Discontinuous variation. 



these plants, the resulting curve will be approximately smooth. 

 Other and different factors may influence the rate or the form 

 of development, hence the size of the young plants if they are 

 grown from seeds sown broadcast in a field. Some may receive 

 more moisture, others will have more shade, or more room than 

 the average. The result will always be, that the variation in 

 such a group is continuous, and that the variation of the result- 

 ing plant, in respect to such a character as height, when ex- 

 pressed graphically, will yield a typically normal Quetelet's 

 curve. It is variations such as these that are termed fluctu- 

 ations. Ordinarily the effect of a variation in the non-inherited 

 developmental factors will be a continuous, fluctuating varia- 

 bility. But this is not always true. It may be that a non-inher- 

 ited developmental factor varies continually, but that the 

 effect of this variation is not an equally continuous variation. 

 For instance, in the relation between such a factor and the 



