32 On the Distribution of Temperature in the Gulf Stream. 
Not to multiply diagrams too much, I have omitted those for 
Hatteras and Cape Fear, which, besides, have nothing very espe- 
cially characteristic in them, unless it may be that the Hatteras 
section is one where the results are most disturbe 
so next pee te No. 7, shows the results of the Charleston 
section. The groups m, n, ‘and o, from 5 and 10 fathoms, 20 and 
30 fashound, a 20, 30, and 50 fathoms, resemble each other very 
much; p, the mean of 70, 100, and 150 fathoms, is slightly irreg- 
ular; and so is r, the 400 fathoms curve. The curves from 5 to 
50 fathoms show extremely well the division of the stream, the 
first or ‘cold wall” minimum, the first, second, and third maxima, 
2298 the second ovis third minima. In-shore, from the first mini- 
mum, isamaximum. ‘The irregularities in these observations, 
though obtained hh registering instruments at such considerable 
depths, are less than those which stationary thermometers, sunk 
in the — show in the’passage of heat to them 
i o. 8 represents the Canaveral section, where the 
same voile thick have been stated are shown, but on a very 
diminished scale. The observations were not carried far enough 
from the shore to reach the second maximum. — first and 
ell 
I have omitted, for reasons already stated, the one of the 
other sections. The same phenomena are in general repeated i ing 8 
all the sections. © a 
The permanency of the division of the stream in different 
years, and the accuracy and sufficiency of the observations, may ._ 
be tested in two ways: the first by comparing the results of run- 
ning the same section in different years by different observers; 
the second by the consistency or inconsistency of the results ob- 
tained at/one depth when compared with those at other depths. 
In order to compare the results of different years, some one sec- 
tion was to be explored in the successive seasons. Thus the 
Cape Henry section, connected the work of Lieuts. G. M. Bache, 
S. P. Lee, and R. Bache, having been explored by each. The 
Hatteras section was common to the work of Lieuts. R. Bache 
and J. N. Maffitt, and the same Charleston section was intended 
to be run by Lieuts. Maffitt and Craven 
he Cape Henry section was three times run over; and it 
appears, by comparing the results of each season with the mean 
of the whole, t that in each group of observations represented by 
one of the curves of diagram No. 6, there is an uncertainty of 
rather less than seven miles in the determination of the maxima 
and minima generally. The best determined points are the first 
and second minima and maxima—the “cold wall” minimum and 
axis maximum having an average probable error of 54 miles, 
the other three points have an average probable error of eight 
miles. 
