326 



SCIENCE 



[N. S. Vol. XXX. No. 767 



and complication, was continual and per- 

 sistent until the climax was reached and 

 the extreme forms died out. At the same 

 time, although the paleontologist must re- 

 gard this as a natural and normal phenom- 

 enon not directly correlated with the habits 

 of the race of animals in which it occurs, 

 and although he does not agree with the 

 oft-repeated statement that deer may have 

 "perfected" their antlers through the 

 survival of those individuals which could 

 fight most effectively, there may neverthe- 

 less be some truth in the idea that the 

 growths originally began where the head 

 was subject to irritating impacts and that 

 they so happened to become of utility. 

 Fossils merely prove that such skeletal out- 

 growths appear over and over again in the 

 prime and approaching old age of races; 

 they can suggest no reason for the particu- 

 lar positions and shapes these outgrowths 

 assume in each species of animal. 



It appears, indeed, that when some part 

 of an animal (whether an excrescence or a 

 normal structure) began to grow relatively 

 large in successive generations during 

 geological time, it often acquired some 

 mysterious impetus by which it continued 

 to increase long after it had reached the 

 serviceable limit. The unwieldy antlers of 

 the extinct Sedgwick's deer and Irish deer 

 just mentioned, for example, must have 

 been impediments rather than useful 

 weapons. The excessive enlargement of the 

 upper canine teeth in the so-called saber- 

 toothed tigers {Machcerodus and its allies) 

 must also eventually have hindered rather 

 than aided the capture and eating of prey. 

 The curious gradual elongation of the face 

 in the Oligocene and Miocene mastodons, 

 which has lately been described by Dr. An- 

 drews, can only be regarded as another il- 

 lustration of the same phenomenon. In 

 successive generations of these animals the 

 limbs seem to have grown continually 

 longer, while the neck remained short, so 



that the head necessarily became more and 

 more elongated to crop the vegetation on 

 the ground. A limit of mechanical in- 

 efficiency was eventually reached, and then 

 there survived only those members of the 

 group in which the attenuated mandible 

 became shortened up, leaving the modified 

 face to act as a "proboscis." The ele- 

 phants thus arose as a kind of after- 

 thought from a group of quadrupeds that 

 were rapidly approaching their doom. 



The end of real progress in a developing 

 race of backboned animals is also often 

 marked by the loss of the teeth. A regular 

 and complete set of teeth is always present 

 at the commencement, but it frequently be- 

 gins to lack successors in animals which 

 have reached the limit of their evolution, 

 and then it soon disappears. Tortoises, for 

 instance, have been toothless since the 

 Triassic period, when they had assumed all 

 their essential features; and birds have 

 been toothless since the end of Cretaceous 

 times. The monotreme mammals of Aus- 

 tralasia, which are really a survival from 

 the Jurassic period, are also toothless. 

 Some of the latest Ichthyosaurs and 

 Pterodactyls were almost or quite toothless ; 

 and I have seen a jaw of an Upper Cre- 

 taceous carnivorous Dinosaur {Genyo- 

 dectes) from Patagonia so completely desti- 

 tute of successional teeth that it seems 

 likely some of these land reptiles nearly 

 arrived at the same condition. 



Among fishes there is often observable 

 still another sign of racial old age- 

 namely, their degeneration into eel-shaped 

 forms. The Dipnoan fiishes afford a strik- 

 ing illustration, beginning with the nor- 

 mally shaped D/p Merits in the Middle De- 

 vonian, and ending in the long-bodied 

 Lepidosiren and Protopterus of the present 

 day. The Paleozoic Acanthodian sharks, 

 as they are traced upwards from their be- 

 ginning in the Lower Devonian to their 

 end in the Permian, also acquire a remark- 



