38 H. M. Parkhurst—Tails of Comets. 
width, it should be in the middle of the tail. On July 17, “v 
was exactly in the center,” and ‘‘its left edge just touched a.” 
(S). These are each within }°. On July 18, “it passed over 
a,” (S.) and @ U. Maj. was within the range of the computed tail. 
“6 and F (g) were not only in the center of the tail but in line 
with its axis.” (S.) They were within 3° of the computed 
center, and very nearly in the computed line of its axis at that 
oint. On July 19, “the tail passed midway between wand f, 
ut touched neither.” (S.) e computed line passes nearly 
centrally between them. “Its left edge just touched 6.” (S.) 
Here is an inaccuracy; for it could not have touched 6 with- 
out passing over 6. On July 20, it passed “midway between 
A and y U. Maj., and centrally over 6.” (S.) This agrees 
til about the 13th, a few days earlier the tail being too short 
for it to be perceptible, and a few days later the curvature it- 
self becoming too small to be perceptible. 
In all these cases the accordance of the right hand edge of 
the tail with the computed line was as close as the nature of the 
observation would admit. But among the observations kindly 
furnished me by Mr. Gilman, the accuracy and faithfulness of 
which I cannot doubt, are two which I cannot in fairness omit. 
On July 3, he recorded the place of the tail as wholly to the 
right of the two stars 65 Arg. 606, 607. The computed line 
passes to the left of those stars. As he did not use a diagonal 
prism, and inverted his chart in making the comparison, it 
seems impossible that he should have put it upon the wrong 
poe especially as his attention was also directed to the star 
65 
nearly a degree further east. On this diagram you will see 
that the observed lines, before and after, are nearly parallel, 
and that these two widely deviate. You will also see, that in 
ch case, connecting the observed point in the tail with the 
position of the head on the previous evening produces a line 
parallel to the rest ; and although Mr. Gilman himself does not 
conceive it to be possible that he connected the position of the 
tail on one day with that of the head upon another, or that he 
could have even seen the comet on July 5, which was Sunday, 
I can find no other satisfactory solution of the discrepancy ; and 
that explanation would substitute two observations accurately 
