4 
150 NATURE 
[Hune 19, 1873 
must be. This conclusion is further corroborated by 
astronomical observations. Science has recorded the 
passage of comets as only some minutes’ distance from 
the surface of the sun ; these bodies must have traversed 
the coronal atmosphere, and yet, notwithstanding the 
lightness of their mass, they did not fall into the sun. 
I shall add here, as to the constitution of the coronal 
atmosphere, a few ideas which do not rigorously flow 
from my observations, but which appear to me very pro- 
bable, but upon which the future must pronounce. 
I said, 2 propos of the observations in the telescope, 
that the corona was shown at Shoolor with a form almost 
square, and that it was distinguished by gigantic dahlia- 
like petals, It is a fact that in each eclipse the figure of 
the corona has often varied ; it has exhibited the most 
eccentric appearances. I have no hesitation in saying 
that this medium, now incontestably recognised, and 
which I propose to name the “coronal atmosphere,” very 
probably does not represent the whole of the aureole 
which is seen during total eclipses. It is quite credible 
that portions of the rings or trains of the cosmical matter 
then become visible and thus tend to complicate the 
figure of the corona, It belongs to future eclipses to 
instruct us on this point. But with regard to the coronal 
medium itself, there is no doubt that it presents singular 
forms, which convey but little idea of an atmosphere in 
equilibrium. Moreover, I am inclined to admit that 
these appearances are produced by trains of very lumi- 
nous and dense matter from the superior layers ploughing 
this troubled medium. The protuberant jets, which 
carry the hydrogen to such great heights, must have a 
peculiar influence upon this coronal medium, whose 
density is quite comparable to that of the cometary 
media. 
It is, then, very probable that the coronal atmosphere, 
like the chromosphere, is very much agitated, and that it 
changes its shape very rapidly, which will explain how it 
presents different appearances every time it has been 
observed. 
To repeat : I have been able to establish at Shoolor, 
by trustworthy and consistent observations, that the solar 
corona presents the optical characteristics of incandes- 
cent hydrogen gas, that this very rare medium extends to 
very variable distances from the sun, from half a radius 
of the sun to about double that at certain points ; but I 
give these figures only as results of an observation, not 
as definitive. It is quite certain, moreover, that the 
height of the corona must be necessarily variable. 
This result seems to be a considerable advance in the 
general problem of the corona. If our foreign rivals 
have not obtained a result so decisive* as those of the 
French mission, I believe it must be attributed to the 
altogether exceptional purity of the sky in the station 
which I chose with such pains, and also to the combined 
optical arrangements which gave to the luminous pheno- 
mena which it was the object to catch, an exceptional ; 
power.t JANSSEN 
CHRONOMETER TESTS 
bi meet following, which has been sent us by the 
Scientific Editor of Harper's Weekly, shows with 
what minuteness the scientific work of this country is 
studied in America, and what a critical audience we have 
on the other side of the water:—One of the most im- 
portant services that astronomy has rendered to man- 
Kind consists in the contributions it has made to the 
* M. Respighi, at Poodookotah, made observations purely spectroscopic 
which confirm mine; only he found the height of the corona much less, 
which appears to me to be due to the more feeble luminous power of his 
instrument. 
_ t This paper contains only an analysis of my obzervaticns: I have not 
been able to refer in detail to those of other observers. I may cite, however, 
he important remarks of Mr. Lockyer on the structure of the corona, the 
pt graphs of Colonel Tennant, the polariscopic observations made at 
Jafna, these of Capt. Fyers, M. Oudemans, and others. 
progress of navigation, and the increased security of life 
and property. In this field England has always taken 
the lead, and the efforts of Mr. Hartnup at Liverpool are 
a worthy continuation of the labours of Flamstead, 
Bradley, and Airy. While the Greenwich Observatory 
has caused a great improvement in the general standard 
of the chronometers bought for the use of the Govern- 
ment vessels, Mr. Hartnup has sought to effect a similar 
reform for the mercantile marine. He has insisted on 
the vital importance to ship-masters, as well as to owners 
and insurance companies, of the careful determination of 
the rates of their chronometers as affected by tempera- 
ture. The makers of these instruments and the astro- 
nomers who use them carefully have always known that 
which captains of vessels have been very slow to profit 
by—z.e. that the chronometers are, when made, so 
adjusted that they keep perfect time at two temperatures, 
such as 55° and 85° F., while between these limits they 
gain, and beyond them they lose, on the true time. It is 
rare that this variation in the chronometer rate can be 
safely overlooked by a careful navigator, though it is 
frequently done by those whose vessels do not carry a 
precious burden of 1,000 or 2,000 souls, The only 
excuse for this neglect is the positive assurance of the 
maker that the chronometer is perfectly reliable—an 
assurance that is often fortified by very deceitful figures. 
The difficulty and expense of a searching investigation 
into the errors to which every chronometer is liable 
have long been supposed by the trade to stand in the 
way of the introduction of such chronometers only as 
were of approved reliability. In order to obviate the 
difficulty as far as possible, the Liverpool Observatory 
has been constructed by Mr. Hartnup specially for the 
purpose of studying the rates of the chronometers that 
may be sent thither by captains sailing from that port. 
The expense of the examination given to such chrono- 
meters is comparatively trifling; and the number of 
chronometers submitted to him has annually increased, 
until by reason of the recent regulations at that port the 
number of examinations has amounted to between 1,000 
and 2,000 annually, the same instruments having been 
repeatedly submitted to him. The process pursued by 
Mr. Hartnup consists in exposing each chronometer for 
a week to a uniform temperature of 55°, and determining 
its rate each day; itis then for another week exposed to 
a temperature of 70°, and then to one of 85°; the next 
week it is returned to the temperature of 70°, and the 
last or fifth week it is exposed to the temperature of 55°, 
as at first. By means of general laws regulating the 
rates of chronometers it is now possible to determine 
what the rate will be at other temperatures than the three 
above mentioned, and knowing these, the navigator is 
able to apply the proper correction to his time-keeper 
so exactly that he need never mistake his position upon 
the ocean. 
The records of the Liverpool Observatory for the past 
year show—1. That the rates of about 10 per cent. of | 
the chronometers tested (those of the mercantile marine 
very generally have the ordinary compensation balance) 
are so irregular as to render the instruments entirely unfit 
for nautical purposes. 2. The error of adjustment for 
temperature of the remaining 90 per cent. is often so 
erroneous as to produce a change of daily rate of many 
seconds, when the temperature varies but little from either 
of the two standard points of 55° and 85°, or thereabouts. 
3. That the best made and most carefully adjusted in- 
struments gain, on the average, daily six-tenths of a 
second more at a temperature of 70° than at 55° or 85°. 
4. That those that have the same rate at 55° and 70°, or 
at 70° and 85°, lose when exposed to temperatures beyond 
these limits at the rate of 1°5 seconds daily for a change 
of 15° in temperature. 5. That when the connection be- 
tween temperature and daily rate has been well deter- 
mined, it will remain constant in good instruments for a 
