64 SEXUAL DIMORPHISM 



the cartilages has directly adapted itself to the pressure 

 produced by the compressed air in the larynx. But this is 

 not all : in the majority of species of these monkeys the front 

 part or the whole of the manubrium sterni, instead of being 

 single and median as it usually is, is divided into two halves. 

 The enlarged parts of the larynx do not lie between the two 

 halves of the manubrium when the animal is at rest ; but 

 when the dilated thyroid and basihyal are expanded by the 

 inflation of the larynx in the production of the prolonged 

 howl, then the upper part of the trachea is forced against the 

 inside of the manubrium and presses its two horns apart. 

 To suppose that a mechanical influence apparently so obvious 

 has no physiological meaning is contrary to reason and 

 common sense. On the view that the mechanical effects 

 of the actions of the body in course of generations become 

 converted by heredity into the regular course of develop- 

 ment, the evolution of the condition is intelligible ; on any 

 other view it is impossible to explain it. The selectionist 

 doubtless supposes that the loudest voiced individuals leave 

 most progeny, and that any variations of structure which 

 increase the power of the voice are thus selected. But there 

 is no reason for the suggestion that such variations as are 

 necessary would ever occur except as a mechanical consequence 

 of the exertion of the voice. The howl seems, like that of the 

 domestic cat, to be sexual, but we have no evidence that it is 

 competitive. The truth seems to be that sexual excitement 

 is expressed in various ways, and that a special habit, once 

 begun, tends to increase by inheritance till it is carried to 

 a maximum, and becomes an instinct. The reason why the 

 howling is practised by the young and by the females may 

 well be that the monkeys imitate one another, as gregarious 

 animals are apt to do. 



In the Cheiroptera or bats sexual differences are not 

 strongly developed, and we may therefore pass on to the 



