May 9, 1884.] 
tied as to suggest such a ‘designation. The choking 
or drying of the cocoons was in colonial days a part 
of silk-raising, and not of silk-reeling; and, while 
reeling-establishments may undertake to choke the 
cocoons brought in by the raisers in their immediate 
neighborhood or by agents, the marketing of fresh 
cocoons must necessarily be limited in time and dis- 
tance. They cannot bear pressure without injury, 
and all baled cocoons must needs be choked. One is 
hardly justified in comparing the methods of colonial 
times with those in vogue to-day in France, where 
modern steam filatures and railroads have produced 
such profound modifications. We cannot see how 
choked cocoons, which have but one-third to one- 
fourth the weight of fresh cocoons, can be marketed 
at the same rates as the fresh cocoons. ‘The term 
‘green’ cocoons is often used in English as the equiv- 
alent of fresh cocoons; but, as quoted in the French 
markets of to-day, the word ‘green’ (vert) refers to 
those of a green or greenish color. Perhaps this may 
explain the puzzle. Co Ve RinEnyY. 
Thermometer exposure. 
In No. 58 of Science, Professor Mendenhall calls 
attention to interesting differences of the minima 
temperatures on cold, still nights of the winter. I 
agree with him that a difference of exposure, and prox- 
imity to buildings, may explain a difference in read- 
ing ; but it is impossible to explain by them alone the 
enormous difference noticed in Columbus (27.8° F.). 
There must have been, besides, one or another of the 
following conditions, probably both. When the con- 
ditions are favorable to radiation, and the night is 
still, the lowest strata of the air are mostly cooled by 
contact with the cold, upper surface of the ground; 
and more so if there is snow, and aso-called inversion. 
of temperature is produced. The temperature rises 
from the lowest strata toa certain height. Examples 
of this can be found in the observations at Pulkova, 
near St. Petersburg. A thermometer placed at the 
height of seventy-eight feet was almost constantly 
higher than one at six feet above ground at eight P.M. 
In August, on clear days, the mean difference was 
2.1° F., and once in September it was 5.2° F. In the 
months from December to March, when the ground 
is covered with snow, even at one P.M. the upper 
thermometer was higher than the lower; the mean 
difference on clear days of December and January at 
one P.M. amounting to 1.3° F., and once it amounted 
to 4.1° F. 
The same results were obtained by experiments 
made at Kew, by direction of the meteorological office. 
The minima were lower at a height of twenty-one feet 
above ground than at a hundred and twenty feet; and 
on one occasion, at nine P.M., during a fog, the latter 
was higher by 10.8° F. than the former. 
Now, most of the signal-service stations must have 
comparatively high minima, not only because they 
are mostly located in the interior of cities, but be- 
cause the thermometers are often placed very high 
above the ground, at the level of the fifth or sixth 
story of city buildings. Probably the stations of the 
Ohio state service are placed lower. 
Besides the height of thermometers above the 
ground, what I call the ‘topographical conditions’ 
are of importance. At an equal distance from the 
level of the ground, under conditions favorable to 
radiation, there will be much lower minima in valleys 
than on hills. This is caused by the descent of the 
coldest and heaviest air to the valley, and also by 
the fact that in a valley the air is in the vicinity of 
a greater surface of the ground. During the anti- 
cyclone of Dec. 19-30, 1879, the summit of Mont 
SCIENCE. 
563 
Verdun, near Lyons, France, had a mean tempera- 
ture of —1.7°.C.; and the Pare de la Téte d’Or, in 
the city, situated four hundred and fifty metres lower, 
a mean of —7.1° C. The mean minima differed by 
more than 12°C. Very likely the observations of 
the state service at Columbus were made on lower 
ground than those of the signal-service. Where anti- 
cyclones in winter are common in high latitudes, with 
the ground covered with snow, the mean temperatures 
of the winter months must be considerably colder 
in valleys than on the surrounding hills and moun- 
tain slopes, as the insolation during the day inter- 
feres but slightly, and not at all during some days at 
points beyond the polar circles, with the equilibrium 
of air strata obtained during the night. 
This cold of the nights in valleys, subjecting plants 
to freezing on nights when those that grow on hills 
are spared, is well known. Perhaps it is less noticed 
in the United States, as there low temperatures are 
oftener accompanied by high winds than in Europe. 
The olive-cultivators in southern France, and the 
coffee-growers in the hilly districts of the province 
of San Paulo, southern Brazil, know this so well that 
they do not plant their trees in valleys, from fear of 
frosts. A. WOEIKOF. 
St. Petersburg. 
Dalmanites in the lower carboniferous rocks- 
During a recent geological excursion near this city, 
one of our party, Mr. Henry Lane, found and pointed 
out to me a trilobite, which J extracted from the 
stone myself. The rock on which we were working 
was the upper part of the Cuyahoga shale of the 
Waverly group of Ohio, now universally, I believe, 
referred to the lower carboniferous system. The only 
genus hitherto reported from these rocks in America is 
Phillipsia, with the exception of two species of Proe- 
tus searcely distinguishable from Phillipsia. The 
specimen in question, however, distinctly differs from 
both of these in the pygidium, the only part yet ob- 
tained. Instead of the evenly rounded and margined 
tail of those genera, it shows the flabellate and fim- 
briate form of Dalmanites. The occurrence of this 
genus or of this type of trilobite, so high in the geo- 
logical series, is both surprising and ‘ uncanoniceal.’ 
EK. W. CLAYPOLE. 
Buchtel college, Akron, O., 
April 14. 
‘A curious optical phenomenon.’ 
Except in one curious point, ‘F. J. S.’s’ latest ex- 
periment (Science, No. 63, p. 475) obviously accords 
with my note (same page). Apparently, the virtual 
image is three feet in front of him, or nine feet from 
the wires, since the phantom rises when he bows; the 
slats are seventeen and two-thirds times wider apart 
than the wires, from centre to centre; and every 
fourth wire hides every third slat, while the next wire 
but one hides a slat-shadow. But how can thirty 
slats and their shadows thus give twelve dark phantom 
lines? With his telescope, ‘F. J. S.’ may find that 
two of them, least perfect, are where wires cross the 
frame of the blind. 
Two words of mine, three lines from the bottom of 
the page, require correction. The size of the image 
is not ‘very nearly’ as described, but exactly so. If 
this image could become an actual screen, then its 
image, in turn, would be the farther screen; and any 
line through a wire-crossing in either of the three 
screens would meet the other two at points quasi- 
homologous to each other. 
. JAMES EDWARD OLIVER. 
Cornell university, April 29. 
