unit of measurement employed in this connection was the stride of my riding-camel. In Pao-t'ou we had measured out a base-line of a couple of hundred meters, and by riding backwards and forwards over this distance several times I had established a mean value for the camel's strides. In the same way I calculated the mean value for the number of minutes necessary to cover this distance. I could thus at any time check our speed by noting how many strides the camel took during the mean value of minutes I had already gained on the base-line. The mean value for the whole day's journey gave the distance. Simple as it is, this method, which I used for many years, gives quite a satisfactory result.