Most of the holes drilled in the stone instruments and pipes of North America appear to have been produced by hollow drills, which Professor Rau[171] suggests may have been formed of a hard and tough cane, the _Arundinaria macrosperma_, which grows abundantly in the southern parts of the United States. He finds reason for supposing that the Indian workmen were acquainted with the ordinary form of drill driven by a pulley and bow. The tubes of steatite, one foot in length, found in some of the minor mounds of the Ohio Valley,[172] must probably have been bored with metal. Dr. Keller, after making some experiments with a hollow bone and quartz-sand, tried a portion of ox-horn, which he found surprisingly more effective, the sand becoming embedded in the horn and acting like a file. He comments on the absence of any bronze tubes that could have been used for boring in this manner, and on the impossibility of making flint tools for the purpose. The perishable nature of ox-horn accounts for its absence in the Lake settlements.[173] On the whole this suggestion appears to me the most reasonable. Experiments have also been made in boring with stag’s-horn.[174] M. Troyon[175] considered that these holes were not bored by means of a hollow cylinder, inasmuch as this would not produce so conical an opening, and he thought that the axe was made to revolve in some sort of lathe, while the boring was effected by |51| means of a bronze tool used in conjunction with sand and water. He mentions some stone axes found in Bohemia, and in the collection of the Baron de Neuberg, at Prague, which have so little space left between the body of the axe and the central cores, that in his opinion they must have been bored by means of a metal point and not of a hollow cylinder. Mortillet[176] thinks that some of the Swiss axes were bored in a similar manner. The small holes for suspension, drilled through some of the Danish celts, he thinks were drilled with a pointed stone.[177] Not having seen the specimens cited by M. Troyon, I am unable to offer any opinion upon them; but it appears to me very doubtful whether anything in character like a lathe was known at the early period to which the perforated axes belong, for were such an appliance in use we should probably find it extended to the manufacture of pottery in the shape of the potter’s wheel, whereas the contemporary pottery is all hand-made. M. Desor,[178] though admitting that a hollow metallic tube would have afforded the best means of drilling these holes, is inclined to refer the axes to a period when the use of metals was unknown. He suggests that thin flakes of flint may have been fastened round a stick and thus used to bore the hole, leaving a solid core in the middle. I do not however think that such a method is practicable. In some of the Swiss[179] specimens in which the boring is incomplete there is a small hole in advance of the larger, so that the section is like that of a trifoliated Gothic arch.