108 



in a recent publication. The significance of these facts could 

 and should be clearly presented in any popular treatise. 



In the review of Mendel's work there is no mention of his 

 report of investigations with beans, in which he found that the 

 apparent unity of characters observed in Pisum did not prevail 

 and in which he suggested what is essentially the multiple factor 

 hypothesis of today. 



Probably no phase of plant-breeding is of greater popular and 

 scientific interest than that of mutation. Among investigators 

 there is much diversity of opinion regarding this, and there is 

 much conflicting evidence which would hardly be suspected by 

 the treatment in the volume under consideration. In the first 

 place, doctrines of mutation were based on the simple conception 

 of unit-characters and have not been brought into close harmony 

 with the revisions of that conception. Furthermore, de Vries 

 does not consider that the loss of hereditary factors or characters 

 is usually associated with a mutation. The hereditary units he 

 conceives as either stable or labile. When labile they may 

 change from active to inactive or latent, or to a semi-latent con- 

 dition. It is not the presence or absence, the gain (with the ex- 

 ception of the few progressive mutations) or the loss of well- 

 defined units as most Mendelians take for granted, but the 

 varying degrees of activity of ever-present "pangens" that 

 bring about mutations. This view of de Vries is not even 

 considered by the writers (see page 193). 



The treatment of graft-hybrids is decidedly inadequate if not 

 misleading. The facts recently developed in connection with 

 studies of graft-chimeras and graft-hybrids are undoubtedly of 

 greater significance in their bearing on the fundamental principles 

 of heredity than are those of any other line of investigation 

 developed during the last decade. The excellent studies of the 

 cell relations in Cytisus Adami in the Crataegus-Mespiliis chimeras 

 and in Solanum tubingense, S. proteus, S. Koelreuterianum, S. 

 Gaertnerianum which are already treated in standard texts like 

 Jost's Pflanzenphysiologie and the Strasburger text-book show 

 that all these are periclinal chimeras. A few words regarding 

 the cell relations in these plants would make clear their nature 



