[ 54i ] 
caufed the line to rub againft the Safeguard hole, for 
the button to flop againft, as above defcribed ; which 
undoubtedly occalioned confiderable fridfion in that 
place. But the moft untoward circumftance, that I 
obferved, was, that being in a rough, but Ihort chop- 
ing fea, and failing obliquely acrofs the waves, the 
plate would frequently be drawn from one wave to 
another through the air, without touching the water ; 
and, as it appeared, would jump from one wave to 
another, the unevennefs of the furface, joined to the 
quicknefs of the motion, not permitting the plate to 
follow the depteffion of the water. This evil I en- 
deavoured to remedy, by placing upon the line, at a 
fmall diltance before the plate, fome hollow bullets, 
fuch as are made for nets, in order to keep the plate 
fo low down in the water, as to be below the bottom 
of the waves. This, in part, 1 found they did ; but 
they, at the fame time, added fo much refinance, in 
their palling through the water, that the inconvenience 
was as great one way, as the other. 
Upon making up the account of this run, I found 
the number of rotations were lefs, by one full third, 
than they ought to have been, compared with former 
obfervations ; which afforded me a convincing proof, 
that this inftrument was conliderably retarded in quick 
motions 
The length of the line made ufe of was about 10 
fathoms, which I found necefftry, that the water, 
difturbed by the body of the fhip, might be tolerably 
fettled before the plate was drawn through it ; but 
this length of line was alfo an inconvenience,, as it 
met with greater refiftance in the water. 
