August 14, 1696.] 



SCIENCE. 



185 



widely separated pliyla, such as mammals, 

 gasteropods and cephalopods, are so con- 

 sistent and harmonious as to be full of 

 promise for the future. 



Limitations of time and space forbid an 

 attempt to fully consider here all the de- 

 ductions which have been suggested and 

 rendered more or less probable by this 

 method, but one or two principles which 

 stand out with especial clearness may be 

 mentioned. 



(a) Evolution is ordinarily a continuous 

 process of change by means of small grada- 

 tions. The continuous character of a 

 phylum is apt to be proportional to the rela- 

 tive abundance of its representatives in the 

 strata, which is equivalent to saying that 

 well-known series are continuous, while ap- 

 parently discontinuous series are imper- 

 fectly known. This does not imply that 

 the rate of change was always uniform — it 

 probably was not — or that a sudden alter- 

 ation of conditions may not bring about 

 discontinuity, or ^er saltum development. 

 It means that the usual and normal mode 

 of advance is by continuity of change. 



(6) Development is, in most instances, 

 direct and unswerving. The rise of new 

 forms, and the decadence and degeneration 

 of old ones, are not ordinarily by zigzag and 

 meandering paths, but by relatively straight 

 ones ; and though, of course, a path once 

 taken may be diverged from, yet in such a 

 case it is not regained. This applies par- 

 ticularly to the organism as a whole ; in 

 minor details more latitude is permissible. 

 The evidence is not yet sufficient to show 

 just how widely applicable this principle is. 



(c) Parallelism and convergence of de- 

 velopment are much more general and im- 

 portant modes of evolution than is commonly 

 supposed. By parallelism is meant the 

 independent acquisition of similar structure 

 in forms which are themselves nearly re- 

 lated, and by convergence such acquisition 

 in forms which are not closely related, and 



thus in one or more respects come to be 

 more nearly alike than were their ancestors. 

 While some observers have tacitly or ex- 

 plicitly denied the reality of these processes, 

 most authorities have been compelled to 

 admit them. What paleontology has done, 

 and is doing, is to show the universality of 

 these modes of development, and to point 

 them out in directions where they had not 

 been suspected. To give a few examples : 

 The crescentic, or selenodont, molar has 

 been separately acquired by no less than 

 three groups of artiodactyls, and probably 

 others as well. The spout-shaped odontoid 

 process of the axis has independently de- 

 veloped in the horses, the tapirs, and in 

 three artiodactyl series. The true rumi- 

 nants (Pecora) of the present day are, 

 among other characteristics, distinguished 

 from the remaining artiodactyls b}' the hol- 

 low tympanic bullae, which in the pigs, 

 tragulines and camels are filled with can- 

 celli, or spongy bone. In Oligocene times 

 only the camels had acquired the cancelli ; 

 the other groups, though already differen- 

 tiated as such, still had hollow and inflated 

 tympanies. Lists of such parallelisms in 

 single characters might be multiplied almost 

 indefinitely, but they also occur in whole 

 groups of structures. The camels have in 

 teeth, skull, vertebrse and limbs many points 

 of resemblance to the true ruminants, which 

 demonstrably are not due to inheritance 

 from a common ancestor. The two great 

 series of ungulates, the artiodactyls and 

 perissodactyls, which are usually grouped 

 together as the Ungulata par excellence, are 

 examples of parallel development on a grand 

 scale, their many resemblances being for the 

 most part independently acquired. The 

 flesh eaters known as Carnivora include at 

 least two, and probably three lines, which 

 have been separately given off from the 

 primitive flesh eaters, or creodonts. 



Such a mode of development greatly in- 

 creases the difficulty of determining phy- 



