289 
from Iceland to Greenland. The fact that this was effected 
before the time of Columbus proves that it was not improbable 
even in an earlier age. This probability is strengthened by 
the fact that an inscription of an apparent Runic character 
was discovered some years ago on an island off the coast of 
Maine. Another remarkable inscription was found among 
the excavations of Grave Creek, engraved on sandstone, con- 
taining at least four or five Runic letters. As late, also, as 
1859, a stone axe was ploughed up in New Jersey, having 
apparently Runic characters carefully inscribed upon it — 
known among the American ethnologists as the Pemberton 
axe. Facts such as these cannot be overlooked. Possibly, 
when further discoveries shall have increased our knowledge 
by accumulated facts, we may be able to speak upon the sub- 
ject with greater confidence. 
What I have attempted to show, thus far, is that an 
entrance into America, both from Asia and Europe, was not 
only possible, but probable, even in prehistoric times. In 
other words, the unity of the native American races with 
those of the Old W orld, considered as a question of descent 
and of common origin, not only finds no insuperable dif- 
ficulty from the oceanic isolation of the great Transatlantic 
continent, but is almost suggested to us by a variety of 
well-ascertained facts. 
II. In coming to the next subject, viz., the analogies which 
may be traced between the languages of America and the Old 
World, it will be necessary to proceed with the greatest 
caution ; for it is not by discovering a word here and a word 
there seeming to agree with corresponding roots among all 
these various families that we can safely infer any true 
affinity between them. To speak on this subject with 
prudence and wisdom, it will be needful to observe two most 
rigid rules : 1st., To take no notice of any apparent identity 
among words which are evidently based upon imitative 
sounds. 2ndly, Never to deduce an affinity between any of 
the New and Old World languages on the ground only of root 
words existing in common between them ; unless, indeed, the 
same roots appear and reappear, in many distinct tongues 
belonging to large family groups on both sides. In that 
case, I think, the widespread distribution of common root 
words can scarcely be referred to accidental resemblance; 
still less can they have arisen from recent intercourse. In all 
other cases, however, it will be necessary to go beyond such 
analogies. Hence I add (3rdly), Never to deduce an affinity 
in such cases between any of the New and Old World 
family stocks, unless there be a certain intercommunity of 
