107 



of plant-breeding subjects, and also a series of twenty-seven 

 laboratory exercises for class instruction in plant-breeding. 



During the nine years that have elapsed since the appearance 

 of the fourth edition, there has been much investigation, espe- 

 cially with reference to the experimental phases of hybridization. 

 To summarize adequately, critically and impartially these results 

 is not a simple task, yet it is one which justifies the revision in 

 question, and it is in this respect that the new volume is most 

 defective. 



For example, it is stated and in general maintained that 

 characters such as the "presence or absence of pubescence on the 

 leaves, the height of the plant, whether dwarf or tall, the color 

 of the flower or fruit" are unit characters of which plants and 

 animals are composed (p. 9). This we may note was the view 

 held several years ago. Recent investigations and critical 

 studies of the older investigations show that characters which 

 behave as consistent units in hybridization are indeed rare. 



The authors are fully aware (as intimated especially on pages 

 viii, 128, 179, 185, and in Chap. Ill) that the older conceptions of 

 the unity of characters, of dominance, of segregation into parental 

 characters only, and of the purity of the germ cells formed by a 

 hybrid have all been modified by numerous subsidiary hypotheses 

 which attempt to account for increased variability, unexpected 

 ratios and the appearance of intermediate characters. In fact, 

 characters are now considered to be so complex in heredity that 

 even the most enthusiastic Mendelians have discarded the term 

 "unit-character" and substituted the rather intangible term 

 *" unit-factor." 



It is clearly pointed out in the fourth edition (p. 166) and 

 reiterated in the new edition (p. 168), in harmony with Bateson's 

 views some ten years ago, that the most important and crucial 

 point of Mendelian doctrine pertains to the assumed purity of 

 the germ cells produced by hybrids with respect to pairs of 

 contrasting characters. At the present time the evidence not 

 only indicates that few if any characters are in any sense con- 

 tinuous units in heredity, but that also there may be all degrees 

 of impurity in segregation, a condition fully admitted by Bateson 



