299 
neighbouring country ; but on these points all local knowledge 
has perished. 
If popular recollection of distant events is very imperfect, even 
when it is aided by the existence of an historical literature, it is 
a much more uncertain vehicle for the transmission of facts, 
when it is forced to rely on its own unaided resources. In fact, 
events transmitted orally become speedily varied, coloured, and 
exaggerated. This is particularly the case with respect to 
numbers, even when the events are recent. I can well recollect 
the surprise with which I first learned the numbers which were 
engaged at Waterloo, compared with the popular exaggerations 
of them. We may lay it down as a general rule that popular 
conceptions of numbers are nearly always exaggerated, and when 
handed down to us by mere tradition grossly so. Hence, the high 
numbers so generally found in ancient writers* When we take 
into consideration that the hosts of Xerxes, after they had passed 
the Straits of Thermopylse, could have derived their subsistence 
only from supplies which must have been transported by sea, it 
is evident that the accounts which have been handed down 
as to the numbers of the army and the camp-followers are 
unworthy of credit. In fact, the mode in which they were said 
to have been ascertained was the roughest possible. The late 
war proves that the numbers of armies on paper and of those which 
took the field differ widely. Ancient writers have given the num- 
bers of the Persian force which fought at Marathon as varying 
from 100,000 to 600,000 men. We have a solid fact by which to 
test the truth of this report. The whole was conveyed across 
the iEgean in 600 trireme galleys, the ordinary crew of oue of 
which consisted of 200 sailors and thirty marines. For these 
the space on board was so limited, that whenever a favourable 
opportunity presented itself, they took their meals on land. You 
are aware that the accounts handed down of the earliest portions 
of Roman history are filled with minute specifications of num- 
bers. If these accounts of the numbers which fell in battle are 
worthy of credit, the inhabitants of that portion of Italy must 
have been more prolific than mice. One army is no sooner 
slaughtered that another is in the field, and this year after year. 
But it will be more satisfactory to test the value of oral 
tradition as an accurate reporter of events, not through the 
remote past, but by the recollections of the times in which we 
live. Let us take an instance very favourable for the trans- 
mission of traditionary historical recollections, — the inhabitants 
of a great naval port. Everything in such a place would tend to 
keep alive the knowledge of events, the esprit de corps of a 
constant succession of seamen, the interest felt by the whole 
population in their actions, aud the ships which would help to 
