Thus in the island of Kiwai, at the mouth of the Fly River in New Guinea, the Cambridge Expedition found a whole set of phrases in vogue, whereby the number of subjects acting on the number of objects at a given moment could be concretely specified.

To indicate the action of two on many in the past, they said _rudo_, in the present _durudo_; of many on many in the past _rumo_, in the present _durumo_; of two on two in the past, _amarudo_, in the present _amadurudo_; of many on two in the past _amarumo_; of many on three in the past _ibidurumo_, of many on three in the present _ibidurudo_; of three on two in the present, _amabidurumo_, of three on two in the past, _amabirumo_, and so on.

Meanwhile, words to serve the purpose of pure counting are all the scarcer because hands and feet supply in themselves an excellent means not only of calculating, but likewise of communicating, a number.

It is the one case in which gesture-language can claim something like an independent status by the side of speech.
For the rest, it does not follow that the mind fails to appreciate numerical relations, because the tongue halts in the matter of symbolizing them abstractly.

A certain high official, when presiding over the Indian census, was informed by a subordinate that it was impossible to elicit from a certain jungle tribe any account of the number of their huts, for the simple and sufficient reason that they could not count above three.