364 RAYMOND PEARL. 



case of peas as are here shown to exist in maize. That such 

 variation, provided it be really somatic or fluctuational, is, how- 

 ever, of no real importance in relation to the cardinal facts of 

 Mendclian inheritance has been shown by all experimentalists 

 who have devoted attention to the matter. Bateson 1 (loc. cit., 

 pp. 240-244) gives an illuminating discussion of the whole matter, 

 with special reference to the phenomena in peas. East and 

 Hayes 2 discuss the same point with reference to maize and sh<>\\ 

 that somatic intermediates behave in inheritance in accord with 

 their gametic constitution rather than their somatic appearance. 

 Certainly the time is past when facts such as are set forth in the 

 present paper can be adduced in criticism of basic Mendelian 

 principles. 



The essential point brought out by this study is, it seems to me, 

 that the well known general fact that every datum of science 

 is a function (in the mathematical sense) of two variables, namely, 

 the observer and the thing observed, is once more emphasized 

 by a particular case. 



A thorough investigation which brings out essentially this 

 same point, though conducted on a different class of material 

 and with a somewhat different object in view, has been made 

 by Yule. 3 



It will be freely admitted by everyone as an abstract proposi- 

 tion that the personal idiosyncracy of the observer constitutes a 

 source of error in all scientific observing. Yet how often does 

 the biologist not working on strictly quantitative problems make 

 any effort either to eliminate or determine the magnitude of this 

 source of error in his case and in a specific instance? Anyone 

 who has not experimented for himself on the matter can hardly 

 realize how important, on the one hand, and how difficult on tin- 

 other hand, it is to attain to any considerable degree of real 

 objectivity in results. While the "exact" sciences are some- 

 what better off in this regard than biology, they are after all not 

 greatly so. There has, to be sure, been a great deal of work d< >IH- 



1 Bateson, W., "Mendel's Principles of Heredity." 2cl Edit., Cambridge, 1909. 



2 East, E. M., and Hayes, H. K., "Inheritance in Maize." Conn. Agr. K\pt. 

 Stat., Bulletin 167, 1911. 



8 Yule, G. U., "On the Influence of Bias and Personal Ki|ii;itinii in Statist!' 

 Ill-defined Qualities." Jour. Anthropol. Insl., Vol. XXXVI.. pp. .i-'.S Ji, 1906. 



