150 BOTANICAL GAZETTE [ArcusT 



hitherto regarded as independent of one another, as the association of the color 

 and hairs of the lea\^s of the seedlings of common stock with double flowers; 

 of the characters of the foliage of the quince "with the qualities of the fruit; of the 

 form and hairiness of the scales of barley with the practical qualities of the grain. 

 The author believes that a discover}^ of the laws governing such ^^correlations" 

 may give us a certain amount of power over them that will prove of immense 

 practical and experimental importance. The second section illustrates ''Correla- 

 tions in agricultural breeding," and it is stated that ^'to the practical breeder it 

 shows the way in nearly all the burning questions, and for the scientist it may 

 give the solution of numerous problems which have eluded his evolutionary specu- 

 lations for more than half a century." The third section, entitled "A methodi- 

 cal study of correlations," describes in detail the remarkable system of records 

 developed at Svalof to keep track of all the marks and their associations. It is 

 called "a model of the combination of science and practice." The fourth section 

 deals with "Correlations in fluctuating variability," and is a theoretical discus- 

 sion of the influence of external life -conditions on the phenomena of correlation, 

 illustrations being taken from teratology and from agricultural plants. The 

 cause of this parallelism between different organs is seen in nutritive factors most 

 prominently, minor ones being temperature, moisture, light, etc.; and illustrations 

 are given to show how such factors would affect several regions simultaneously. 

 The last section deals with "Unit-characters," and is a most fundamental specula- 

 tion from the standpoint of evolution. Each organism is conceived of as a "micro- 

 cosm, consisting of thousands of elementary entities, which combine to give 

 it its form and function." These are the units which govern and control the 

 visible characters and qualities. The larger the number of common umts and the 

 smaller the number of different units, the greater will be the affinity. As to the 

 nature of a "unit," there are two points of view. They may be approached by 

 analyzing the visible characters and reducing them to independent groups; or 

 there may be some invisible, although material cause, which constitutes the real 

 source of each unit. There is no reason to assume that a unit should be limited 

 to one organ, to one tissue, or to one cell. A imit may show its acti\aty in different 

 organs, sometimes even in almost all parts of a plant. This explains correlations 

 and "has overwhelming importance in hybridism." Crosses give insight into 

 the nature of imit-characters, showing what marks belong together, and dividing 

 so-called characters into their constituent units. When one unit is added to or 

 subtracted from a well-known type, the result is a mutation. The author is 

 impressed with the idea that the study of correlations must be conducted on the 

 broadest possible lines^ and that in this direction we shall approach some definite 

 knowledge of evolution. 



The last topic is "The geographical distribution of plants" and is chiefly 

 an expression of doubt as to the real value of the current ideas concerning the 



,daptations 



For example, the current 



view of desert plants is that they are "astonishingly specialized and adapted for 



impossi 



" It is a natural 



