when driven into different species of Timber. 39 
_ The first five of the preceding experiments show that with a spike 
of given form, and driven a certain distance into different timbers, 
the order of retentiveness, beginning with the highest, is as follows : 
1, locust; 2, white oak; 3, hemlock ; 4, unseasoned chestnut ; 5, 
yellow pine. From the 6th, 7th and 8th experiments, we see that 
chestnut is still above yellow pine, but that hemlock is inferior to 
both. By the 9th and 10th, it also appears that hemlock is still to 
be placed below chestnut. Comparing the Ist experiment in this 
table with the 6th, and the 2nd with the 7th, we perceive that the 
swell towards the point of the spike, was so far from being an ad- 
vantage to it, that it in fact rendered the spike less retentive than 
when that swelled part had been removed; so that, even could this 
form have been produced without any increase in the weight of the 
spike, it would still have been less advantageous than the simple 
groove without the swell: but when it is considered that the swell 
added 47 grs. (=806 -759) to the weight, it is evident that the 
groove alone has a decided advantage over the other form. By 
the trials in unseasoned chestnut, (Nos. 1 and 6.) this advantage is 
2440 —2121 
15 per cent. ; thus 9a, = 15; and by those on yellow 
2328 — 2069 
pine, (Nos. 2 and 7,) it is 2069 = 12-5 per cent. In fact, 
afier the ends of the fibres have once been thrust apart by the thick 
part of the swell, it is evident that when they come opposite to the 
cavity above the swell they must lose some portion of their power to 
press the spike and produce the retaining force of friction ; this force 
must then depend for its production on the action of those fibres of 
the wood which are opposite to the swelled portion, or between it 
and the point of the spikes. 
In the next series of experiments, it was attempted to ascertain 
the relation between forms ‘more diversified than had hitherto been 
employed. 
As it is evident that the total retentiveness of the wood must de- 
pend, in a considerable degree, upon the number of fibres which are 
longitudinally compressed by the spike, it was inferred, that on the 
area of the two faces, which in driving the spike are placed against 
the ends of the fibres, must in a great measure depend the retention 
of the spike. In this series, four kinds of wood and ten forms of 
spikes were employed. 
