April 7, 1905.] 



SCIENCE. 



531 



than a truism to say that if an anomaly is 

 to be explained by reversion the structure 

 in question must have been normal in some 

 ancestor; but in view of the vast number 

 of anomalies, we are forced to believe that 

 this ancestor must have been a museum of 

 anatomical curios of the most diverse na- 

 tures. It is overlooked that if the explana- 

 tion of a reversion be true it must apply 

 not to one only, but to every possible devia- 

 tion of structure that is not pathological. 

 Phylogeny must show nothing in the his- 

 tory of the supracondyloid process, for in- 

 stance, which will not accord with that of 

 the paroccipital process, or with that of 

 the third trochanter, or with that of each 

 and all of the hundreds of variations which 

 the human body may present. Not only 

 has this accord not been shown, but obvious 

 contradictions have been neglected. An 

 explanation has been sought by referring 

 certain peculiarities very far back : even to 

 a hypothetical common vertebrate stem an- 

 tedating the classes. We admire the learn- 

 ing and the research ; but does the explana- 

 tion explain? 



One of the great difficulties of selection 

 has been to account for the appearance of 

 strikingly similar adaptations or arrange- 

 ments in species from entirely different 

 lines of descent. Analogous to this is the 

 similar irregular appearance of variations. 

 The fossa praenasalis is a deep, sharply 

 marked depression just below the nasal 

 opening, occurring chiefly in low races. It 

 is not to be confounded with the gradual 

 passage of nose into face which is the rule 

 among mammals. I am not aware that it 

 is found among mammals except in the 

 seal, and even there it is less well defined 

 than it may be in man. Here, then, is a 

 sudden change not atavistic and certainly 

 not progressive. The pronator quadratvis 

 in man very rarely sends a prolongation 

 to a carpal bone. I have found this as a 

 variation in a chimpanzee, and Macalister 



in a lion, but in no mammal is it normal. 

 To find it also in turtles and in the Crypto- 

 branchus japonicus does not help us much 

 towards an explanation. 



It is very suggestive that in certain vari- 

 ations of the platysma by which it enters 

 into various combinations with the facial 

 muscles, in some cases its fibers are in direct 

 continuity with those of muscles which 

 comparative anatomy teaches belong to an- 

 other layer. It seems as if nature were 

 striving for a certain effect and is abso- 

 lutely indifferent by just what means it is 

 accomplished. 



One of the most significant points of the 

 mutation theory is that it rehabilitates spe- 

 cies with its old-time dignity. Though we 

 flounder in our definitions of species, we 

 can not get rid of the thought that it is 

 something, after all. By a strange para- 

 dox it is precisely through variations that 

 the tendency towards stability of species is 

 emphasized. In my observations on the 

 human spine I have found that very fre- 

 quently the effects of a variation in one 

 part are felt in remote parts and, indeed, 

 throughout the spine. Some of these seem 

 directly teleulogical, others tend to preserve 

 the type. Thus if the last ribs are very 

 small, and this holds good whether they be 

 the normal twelfth pair or the abnormal 

 thirteenth, the rib before the last is usually 

 exceptionally long. In the case of cervical 

 ribs it is common to find the last rib very 

 small, as if the whole thorax had moved 

 up. In cases w^here there are only eleven 

 thoracic vertebrge not rarely an increase of 

 their size tends to preserve the proper pro- 

 portions of the thorax. In the lumbar re- 

 gion there are certain striking character- 

 istics in the spread and in the structure of 

 the last three transverse processes which 

 give a definite shape to the whole region. 

 In many cases of numerical variation there 

 is an evident effort to reestablish normal 



