572 
NATURE 
[ Oct. 14, 1886 
the anterior of these to the subopercle. Another spot exists at 
the shoulder above the gill-opening, and an indistinct one on 
the edge of the preopercle. 
The figure in the magnificent work of the ‘‘ Fishery Industries 
of the United States,” plate 202, agrees with this specimen in 
the proportions of the body and length of the fin-rays ; but, sup- 
posing the species to be the same, differs from it materially in 
some other points. The position of the dorsal fin is shown too 
far forward; the scaling is not identical, as the lateral line 
in Capt. Gray’s specimen stands upon a narrow but vertically 
elongated row of smooth scales, having three rows of smaller 
ones above it, and four below it; also the eye-like spots are 
not similarly placed, and the pectoral fin is narrower, with its 
rays more elongated. FRANCIS Day 
Cheltenham, October 1 
The Sense of Smell 
WITH reference to Mr. Mitchell’s inquiries (NATURE, Sep- 
tember 30, p. 521), there is a peculiarity about musk which I 
have never found anybody to be previously aware of, namely, it 
is impossible to smell it ¢w7ce, taking two good “sniffs” con- 
secutively at a plant, z.e. after a single exspiration; on the 
second inspiration there is no odour of musk whatever. 
GEORGE HENSLOW 
Humming in the Air caused by Insects 
Your correspondent who writes on the above subject in this 
week’s number of NATURE (p. 547) remarks that ‘‘It is singular 
that no explanation has been offered by any one for such a 
common phenomenon.” May I be allowed to refer him to my 
‘* Observations in Natural History” (published in 1846), p. 226, 
where I have given a statement of my own on the subject, 
adding a reference to Humboldt’s ‘‘ Personal Narrative,” in 
which he makes some remarks on this humming, as heard in the 
tropical regions, where the phenomena is naturally so much 
more striking, and on a wider scale. 
There can be no doubt the explanation of the phenomenon 
given by your correspondent is the correct one. 
Bath, October 9 LEONARD BLOMEFIELD 
THE HONG KONG OBSERVATORY 
R. W. DOBERCK, Government Astronomer at the 
above Observatory, has recently published an 
official report on the astronomical instruments under his 
charge, and on the time service of Hong Kong in 1885, 
the determination of local time being the chief purpose of 
the astronomical branch of the institution. 
The Report states that the Observatory possesses a 
transit instrument, by Messrs. Troughton and Simms, of 
3 inches aperture and 3 feet foca! length. Setting in 
declination is effected by means of two small circles fixed 
on the telescope near the eye-end, and read by levels. 
The axis is perforated for side lamps. The pivots, which 
are made of chilled bell-metal, show no perceptible differ- 
ence between their diameters, but minute irregularities 
appear to exist, though too small to allow their exact 
amount to be determined by means of the axis level. This 
level is used in determining the inclination of the axis, 
and another similar level is provided for use with the 
zenith micrometer in the observation of differences of 
zenith distances on either side of the zenith. 
The eyepiece was originally furnished with one movable 
and seven fixed vertical wires, but the latter after a 
little while began to get entangled with the fixed wires, 
and finally broke. Although it had been found very use- 
fulin the determination of the instrumental constants, it 
was not thought well to replace it for fear lest the per- 
manent wires might become disturbed or broken by it. 
Transits were at first observed over all the seven wires, 
but in 1886 only the five middle wires have been used. 
There are also two horizontal wires about a minute of arc 
apart, and the object the transit of which is to be observed, 
is brought midway between them. ‘The eyepiece and wire 
system can be revolved through a quarter of a revolution, 
so that the vertical wires become horizontal, and can be 
used for determining the differences of zenith distance 
for latitude ; but as the instrument is in constant use as 
a transit, this arrangement has never been made use of, 
there being the less necessity for employing it, as Col. 
Palmer had accurately determined the latitude of the Ob- 
servatory in 1882. 
The telescope rests upon a cast-iron standiwith revers- 
ing apparatus; no change in the inclination has ever 
been perceived to be caused by the reversion. Thestand 
rests on a slab of Portland stone on the top of a brick 
pier sunk 5 feet in the ground, where it is surrounded by 
a brick cylinder to protect it from surface oscillations. 
In 1884 505 transits were observed; in 1885, 313; the 
inclination of the axis was observed 150 times in the 
former and 117 in the latter year. A meridian mark, 
which is viewed through an object-glass of about 66 feet 
focal length is placed about 70 feet to the north of the 
transit instrument ; another meridian mark is 11,354 feet 
to the south across the harbour. 
The standard sidereal and mean time clocks were 
supplied by Messrs. E. Dent and Co. The former has a 
cast-iron back which is firmly screwed to iron bolts 
cemented in the pier placed in the clock-room. The 
pendulum has the zinc and steel compensations originally 
designed for the transit of Venus expeditions. The clock 
was also supplied with a galvanic contact apparatus 
omitting one second each minute, for working a sympa- 
thetic dial in the transit-room, but as the contact-apparatus 
was found to interfere withthe going of the standard- 
clock, its use was discontinued early in 1885, and the 
observations have since been made with a chrono- 
meter which is subsequently compared with the standard- 
clock. 
The mean daily rates during ten-day periods of the 
standard clock are given ina table, and from the rates 
between January 1 and June 9 the following formula for 
the rate at ¢ degrees Fahrenheit is deduced :— 
y = + Is.'247 — 0s.'033 (¢ — 70°). 
The clock stopped twice during the year, viz. on June 12 
and August 23, each time during a thunderstorm. A 
difficulty was experienced in the attempt to determine 
separately the barometric coefficient, since the mean 
height of the barometer in Hong Kong falls regularly as 
the mean temperature rises, but it appeared to be 
insensible. 
It should be noted, however, that the mean rates which 
Dr. Doberck publishes here suggest that the formula 
given above is only correct within certain limits. The 
clock would appear to be compensated for 80° or 85°; at 
least there seems to be no regular variation in its rate 
about these temperatures, the decline in the rate which 
accompanies the rise of temperature up to 80° showing a 
check or even a slight reversal about that point. Prob- 
ably, however, the mean daily rates cannot be trusted te 
the degree of refinement to which they are here given. 
The number of transits observed is decidedly small, and 
the errors of the transit-instrument do not seem to have 
been very frequently or regularly determined during 1885. 
No information is given as to the degree of accordance 
of the daily rates actually observed. 
The mean time clock is similar to the sidereal standard 
clock, and appears to go as well. It is furnished with 
galvanic contact springs, which are pressed every hour at 
the exact second, and send a current through a reversing 
commutator worked by one Léclanché cell, by means of 
which the current that drops the time ball at 1 p.m. is 
closed. 
Some trouble has been experienced with the time ball. 
First the galvanic coil in the mean-time clock-case for 
setting the clock right before the ball is dropped, is not 
strong enough, as it takes nearly an hour and a half to 
correct an error of a second. The lock sent out with the 
