432 PROGRESS OF SCIENCE IN THE CENTURY. 



Brachiopods. The fact is that whenever we settle 

 down to measure, describe, and identify, we find 

 that specific diagnoses are averages; that specific 

 characters require a curve of frequency for their 

 expression; that the living creature is usually a 

 Proteus. There are no doubt long-lived, non-plastic, 

 conservative types, like Lingula, and perhaps a score 

 of other well-known instances, where no visible vari- 

 ability can be proved even in millions of years, but 

 to judge from these as to the march of evolutionary 

 progress is like estimating the rush of a river from 

 the eddies of a sheltered pool. 



In the study of variability it seems possible to dis- 

 tinguish between continuous variation, in which the 

 descendant has a little more or a little less of a given 

 character than the parents had, and discontinuous 

 variation, apparently frequent, in which a new com- 

 bination (say, an elegant vase-like pitcher on a cab- 

 bage leaf) appears suddenly without known grada- 

 tional stages and with no small degree of perfec- 

 tion. Though Lamarck said " Nature is never 

 brusque,'' though we adhere to our statement about 

 the rarity of big Jack-in-the-box phenomena, the evi- 

 dence (e.g., of Bateson) as to the occurrence of dis- 

 continuous variations appears conclusive. Such 

 words as " freaks " and '^ sports '' are open to ob- 

 jection, but they suggest the idea of what Mr. Galton 

 calls " transilient " variations, and the fact that or- 

 ganic structure may pass with seeming abruptness 

 from one form of equilibrium to another. 



It also becomes more and more evident that the 

 living creature in many cases varies as a whole or 

 unity, so that if there is more of one character there 

 is less of another, and so that one change brings an- 

 other in its train. If this be so, we are not restricted 



