ANTHROPOLOGICAL SOCIETY. 87 



which never entirely subsided, but still remains to separate Van 

 Dieman's Land from Astralia. And, even if a Caffre from South 

 Africa were to visit the collection, his attention would be drawn to 

 the same instruments, and he would be able to tell that in this 

 country they were used for the purpose of making loud sounds, and 

 warning the women from the ceremonies attending the initiation of 

 boys. How different the races and languages of Australia and 

 Africa ! yet we have the same use cropping out in connection with 

 the same instrument ; and to complete its history, it must be added 

 that there are passages of Greek literature which show pretty plainly 

 that an instrument quite similar was used in the mysteries of Bacchus. 

 The last point is, that it is a toy well known to country-people, both 

 in Germany and in England. Its English name is the "bull-roarer;" 

 and, when the children play with it in the country villages, it is 

 hardly possible (as I know by experience) to distinguish its sound 

 from the bellowing of an angry bull. 



In endeavoring to ascertain whether the occurrence of the " bull- 

 roarer," in so many regions is to be explained by historical con- 

 nection, or by independent development, we have to take into con- 

 sideration, first, that it is an apparatus so simple as possibly to have 

 been found out many times ; next, that its power of emitting a 

 sound audible at a great distance would suggest to Australians and 

 Caffres alike its usefulness at religious ceremonies from which it was 

 desired to exclude certain persons. Then we are led to another argu- 

 ment, into which I will not enter now, as to the question why women 

 are excluded in the most rigid manner from certain ceremonies. But 

 in any event, if we work it out as a mere question of probabilities, 

 the hypothesis of repeated reinvention under like circumstances can 

 hold its own against the hypothesis of historical connection ; but 

 which explanation is the true one, or whether both are partly true, 

 I have no sufficient means to decide. Such questions as these being 

 around us in every direction, there are only two or three ways 

 known to me in which at preeent students can attack them with 

 any reasonable prospect of success. May I briefly try to state, 

 not so much by precept as by example, what the working of those 

 methods is by which it is possible, at any rate, to make some en- 

 croachments upon the great unsolved problem of anthropology. 



One of the ways in which it is possible to deal with such a group 

 of facts may be called the argument from outlandishness. When 

 a circumstance is so uncommon as to excite surprise, and to lead 



