May 22, 1902] 
NATURE 
77 
imagination to do so, that the sun was exactly in the middle of 
the sail circle, and was forced, after repeated observations, to 
conclude that the sun’s centre appeared to be about the ninth 
part of the distance sr to the west of the true centres. I can 
only offer in explanation of this that it was due to some curious, 
perhaps subjective, effect caused by the part of the arc vysXxw, 
which was very bright, for after this ring had vanished the sun 
certainly appeared at s, without any alteration that I could see 
in the position of the two circles left. 
There was no trace of any luminosity other than that of the 
sky atQ. The brightest part of the whole halo, apart from the 
sun itself and the part of the arc YSX near to the sun, was the 
arc VRW’; but at c, v, Y, X and w, and especially at x and w, 
the light seemed decidedly brighter than in the other parts of 
the rings, though there were no proper ‘‘mock suns.” Since the 
time when I saw this weird and magnificent display, I have often 
tried to deduce the observed curves from the known hexagonal 
forms of ice crystals, and even to produce them empirically by 
reflecting and refracting light from glass models, but so far 
without any success. It will be readily seen that the halo bears 
a certain resemblance to the lunar halo which was the subject of 
a letter to NATURE of May 1 by Prof. Barnard, but the 
two are inreality widely different. Probably the form I have 
just described is of exceedingly rare occurrence, for it presents 
far too grand and curious a spectacle to be visible without ex- 
citing attention, and I have never come across any mention or 
diagram of a halo like it. ee Ge LORTER, 
Eton, May 8. 
Sun-pillar (?) 
Miss HERSCHEL (a careful observer) has just called me out 
to see one. At 7.10 p.m. she saw the sun above a bank of 
clouds, in a somewhat hazy sky, but no clouds above it for a 
space of some 5 degrees. © Above that was a light-fringed belt of 
clouds of great depth. From the sun a parallel-sided pillar of 
light, just like the reflection of the sun in a slightly rippled sea, 
stood upright into, and stopped at, the light-fringe ; it was not so 
bright as the reflection spoken of would be, but markedly 
- brighter than the background sky; colour yellow. Miss Herschel 
had to bicycle home three-quarters of a mile uphill to call me, 
and it was fading before she reached home. 
I was prompt, but too late (7.25) to get a good view. The 
possibility of Martinique dust induces me to send you this. 
Sunset moderately red ; temperature here 42° ; air calm all day ; 
-aith dark sky and damp mistiness. W. J. HERSCHEL. 
Littlemore, May 13. 
Palzolithic Implements in Ipswich, 
ALTHOUGH in isolated cases implements of Paleolithic work- 
manship have occasionally been found in Ipswich, it is only 
during the last few weeks that a deposit containing abundant 
Palzeolithic remains has been discovered. 
On March 21 last, after long searching, I was fortunate 
enough to hit upon this interesting site, and the result has been 
a harvest of implements of very varied types. Mr. Clement 
Reid, whom I acquainted with the discovery, at once came down 
to examine the spot, and under his guidance it will be carefully 
studied. 
The relations of the deposits remain to be worked out, but 
so far show some resemblance to those found at Hoxne and 
Hitchin. 
Among the implements found, some have a thick, ochreous 
patina, while others are almost devoid of it. Most are very 
slightly rolled, but some are still sharp. 
Pointed implements roughly worked at the butt predominate, 
but in one case the butt-end has been carefully sharpened. 
A fine oval implement shows signs of having been worked for 
hafting, as also does a smaller chisel-like form. Implements 
corresponding to those described by Sir John Evans as 
‘crescent like,” a boring implement, a possible sling-stone, 
several ovoid forms flat on one side and raised on the other, 
triangular forms, some thick and heavy, one flat, and a delicate 
leaf-shaped implement, show the variety of purposes which 
these flints were made to serve. 
The position occupied by the Palzeolithic remains appears to 
be that of a silted-up channel cut through Glacial deposits. 
Some of the implements were found at a depth of 12} feet, 
others considerably higher, which may account for the difference 
in their condition. Nina FRANCES Layarb. 
NO. 1699, VOL. 66] 
Brickner’s Cycle and the Variation of Temperature, 
in Europe. 
WE now possess excellent long series of weather-observations 
for many places. It occurred to me lately to apply to several of 
the annual temperature series in Europe an averaging process 
which would tend to bring out the larger waves of variation, or 
at least to show how year-groups of a given magnitude compare 
with one another. I have accordingly considered in groups of 
ten years (I to 10, 2 to II, 3 to 12, and so on) the following 
(see diagram) :—A. Annual mean temperature of Greenwich 
(from 1841). B. That of Geneva (from 1826). C. That of 
Bremen (from 1829). D. That of Vienna (from 1826). 
[The Greenwich curve is drawn on a larger scale than the 
others. The degrees are Fahr., those of the others Cent. The 
position of the curves is simply contrived so that they should 
not cross one another. In the case of Vienna, the continuous 
curve from 1855 is for the Hohe Warte near Vienna; the pre- 
vious dotted curve is approxzmate for the same place, deduced 
from data of the University Observatory in Vienna. The 
Bremen figures used extend only to 1895, the other series to 
1900] 
1825'30 'S ho *S-'So 5 60 °S'Jo 'S 80 “S90 "S$ 1900 
Fic. 1.—a, Greenwich ; B, Geneva; c, Bremen; pb, Vienna. 
Taking the case of Greenwich, it will be understood that the 
first point, for 1845, indicates the average mean temperature of 
the ten years 1841 to 1850, the second that of 1842 to 1851, and 
so on. 
The curves, it will be seen, all agree in showing a long wave 
with crest in the sixties, and extending from a minimum in the 
fifties (Vienna a little earlier) to another about the end of the 
eighties (or beginning of nineties). / 
The Greenwich maximum is reached in 1867, and minima are 
found at, say, 1855 and 1887. The temperatures prior to 1841 
are in some uncertainty; but we should probably be safe in 
saying that an earlier minimum, of magnitude corresponding 
rather to the last, occurred about 18146. Thus we have 1816-55 
= 39 years ; 1855-87 = 32 years ; which agrees very fairly with 
Briickner’s cycle of about 35 years. As to previous maxima, 
we may probably reckon one in the later twenties. The curve 
(with final point in 1895) would appear to be now near a 
second maximum from that date (? 1902); and we might with 
some reason, perhaps, look for another minimum, or con- 
spicuously cold group, in the early twenties of this century. 
The minimum of 1855 for Greenwich is considerably less 
pronounced than that ot 1887, and in Brei.en, too, the earlier 
