COMPARISONS OF Sp, RS, K, ETC. 
773 
mean of eight comparisons Kn=W— 0-0170 grain. In September 1828, by a mean 
of twenty comparisons. Professor Schumacher found K=:Kn+ 0-0198 grain ; whence 
K=U+0-0223 grain. In February and March 1829, by a mean of twenty comparisons, 
Captain Kater found K=W+0-0065; whence K=U+0-0259 grain. This differs 
from the first result, 0-0199 grain (not 0-0299 grain as it is erroneously printed). By 
ninety- two direct comparisons of K with the standard by Captain v. Nehus in June 
and July 1829, K=U+0*0339 grain. In the autumn of 1829 Professor Schumacher 
compared K again with the sum of three brass weights of 5000 grains, 400 grains, 
300 grains and 60 grains of platinum, with which it had been compared on its arrival 
at Altona in 1827, and there was no sensible difference from the first comparison. 
By thirty comparisons in October 1829 and February 1830, Professor Schumacher 
found K=Kn+0-0200 grain. This differs but 0-0002 grain from the result obtained 
in 1828. In April 1844, 2K=Sp+RS+0-0667 grain, Sp + RS=2U— 0-01 1 1 grain. 
Therefore K=U-|-0-0278 grain. 
On taking K out of its case after 1 had received it from Professor Schumacher in 
March 1844, I observed a small fragment of wood, like a grain of coarse sawdust, 
adhering to the under surface of the weight so firmly that I was unable to brush it 
off with a feather, and had some difficulty in removing it with a pointed bit of quill. 
The adhesion of the bit of wood to the weight is due apparently to the pressure pro- 
duced by screwing down very tightly the lid of the box in which it was contained. 
Two or three similar grains were imbedded in the velvet lining of the case. In all 
probability this bit of wood had been attached to K immediately after its first com- 
parison by Captain Kater, when it appeared to be 0-006 grain too heavy, and pre- 
vious to its comparison by Professor Schumacher with the brass weights of 5000 
grains, 400 grains and 300 grains, and the platinum weights of 60 grains. 
By the observations of February and March 1829, K=W+0-0065 grain, and by 
those of June and July 1829, K = U + 0-0339 grain ; whence W= 11+ 0-0274 grain. 
But W=U+0-0194 grain when first compared. Therefore in 1829 W had gained 
0-0080 grain. At the same time RM, which has been very carefully preserved, had 
gained 0-0068 grain. In 1844 the well-preserved troy pounds Ex and L had gained 
0-0089 grain and 0-0146 grain respectively, and Ed and D, which were in a less per- 
fect state of preservation, had gained 0-0221 grain and 0-0226 grain respectively. 
The whole gain of K up to 1844 appears to be 0-0218 grain, about the same as that 
of Ed or D. If, as seems probable, K was compared with W about the end of 1826 
or the beginning of 1827, this error of +0-0218 must include the gain of W up to 
that period. The discordances in the weighings of K may be explained by supposing 
the gain of K, including that of W up to 1829, to be 0-014 grain, and the gain from 
1829 to 1844 to be 0-008 grain, since it maybe assumed that brass having a recently 
polished surface gains weight faster than when its surface is protected by a film of 
oxide ; also, that in the same interval, Sb, which was in some measure protected by 
gilding, had gained rather less than K, and that the bit of wood weighed 0-014 grain. 
5 I 2 
