98 APRIL, 1908
new theory. Observation, experiment, hypothesis, philosophy,
and theory follow each other in rapid succession. Mathemati-
cal seminaries, experimental laboratories, and observations
under natural conditions must all be maintained. The pro-
gress of every branch of science as recorded in the literature
of the last three hundred years shows an instructive series of
failures and successes. The experiments of Alexander Wilson
in 1749 were not repeated until the importance of upper air
exploration was realized and until the students of the modern
weather map perceived that long-range forecasts and even
daily forecasts will never become satisfactory until we fully
understand the upper currents and th6 general circulation of
the atmosphere. It is to the study of this latter subject that
kites and balloons, mountain statione and cloud observations
are now essential, while the interpretation of the results needs
the help of the best mathematical physicists.
It is often stated as a reproof to eminent philosophers that
they are not practical ” that they ‘( know ” but can not ccdo.”
However, in the case of Popof, as of very many meteorologists,
the money needed for practical work was not available and
he could only mark out the methods and the paths for others
to pursue. Fortunate is the “practical man ” who has reliable
theoretical men to guide him in the exploration of nature.
The captain of a vessel would be hopelessly lost at sea if there
was no navigating officer to show the course.-C. A.
IUE UOLUMNS IN GRAVELLY SOIL.
By E D. BOURNE Dated Taylomville, Ky., February B, 1908.
I n the MONTHLY WFATHER REVIEW for October, 1907, is a
notice of an artiole on the formation of ice columns in grav-
elly soil, by Professor Goto, and the statement that. an en-
deavor would be made to get a translation, or abstract of the
same. I have been interested, in an unscientific way, in this sub-
ject for years. About thirty years ago I noticed that occasion-
ally a tiny column would shoot up above the general level of
the group. Upon investigation I found that every one of
these taller columns formed on a seed of horseweed (tall rag-
weed), and always on the end opposite to the germ end.
I have at various times made similar examinations ’ and
always found the same result.
POPOF AND W A N ON THE UBE OF KITES IN ME!l!EOROLOGYp
In 1893 Professor Harrington took up the development of
kite work in the Weather Bureau and during the years 1895,
1896,1897 in successive numbers of the MONTHLY WEATHER
REVIEW we publisht various historical references to those
who advocated or used the kite as a means of sending aloft our meteorological apparatus. We now take pleasure in referring
to still another instance that has lately come to our knowledge
and that is eminently worthy of being added to the record. We allude to a memoir by Prof. A. Popof, of Moscow and
&zany published in Russian in the Journal of the Wnister.of
Public Education for September, 1846, but known to us only
thru an abstract publisht in 1849 by Prof. A. Erman at p.
37d-385 Vol. VI1 of his Archives of Science in Russia. Al-
tho Professor Erman is most widely known by his important
works in terrestrial magnetism yet his interest in climatology
is shown by many articlesin his archives and on every page of his Journey Around the Globe. His profound knowledge
of dynamic meteorology is illustrated by his memoir of Feb-
ruary, 1868, on the general circulation of the atmosphere pub-
h h t in Vol. LXX of the Astronomische Nachrichten. I n the present case Erman, writing in 1849, prefaces his
abstract of Popof’s memoir of 1846, by the remark:
It is to be regretted that the paper kite which in Franklin’s hands brought us such important conclusions as to the electrlcity of the atmos- phere is now scracely noticed by physicists. By glvlng i t a proper size
thls apparatus can, however. be applied with great advantage to the determination of the temperature, the wind directlon, and the quantity of aqueousvaporin the upper etrata of the atmosphere. Indeed for
emall altitudes it has some advantage over balloons, since kites stand for a long time almost immovable so that one can determine the altltude by other mean8 than by angular measurements which take up much time and demand special apparatus. For such altitude determinations
the equation of the curved line formed by the kitegtring seems appro- priate and therefore the mathematical expreselons leading to this end will here be given, and the meteorologists will have to use these in order to determine the altitude of the kite itself or the altitude of any point
on its string. Erman adds that if elastic springs be inserted in the kite line at the
reel and again higher up say at the kite and records be made of the ten-
sions at any moment then a simple formula will give the altitude of the
upper sprlng.
We need hardly repeat the mathematical formulas of Popof,
or Erman’s improvements thereon; they may well be useful
when the kites are not too high and the wind fairly uniform,
but are not adapted to the irregularities of atmospheric cur-
rents and will not give the accuracy demanded in the modern
practise of flying many kites tandem in order to attain the
great altitudes that the Hargrave kite has now brought within
our reach. It is interesting to reflect that if Professor Popof
could have put his ideas into practical execution in Russia in
1846 meteorology might have gained fifty years over its pres-
ent condition. As a rule, however, knowledge progresses
by a system of irregular steps, first an idea, then an experi-
ment; fist the failure of an old theory then the starting of a
FORECASTING ON THE PACIFIC COAST.
By Prof. ALEXANDER G. MCAIJIE. Dated Stan Fnmciw, Cal., Febrtary 4,1908.
In an address delivered before the British Association in
1902, Prof. Arthur Echuster exprest the opinion that (‘me-
teorology might be advanced more rapidly if all routine ob-
servations were stopped for a period of five years, the energy
of observers being concentrated on the discussion of the re-
suits already obtained.” The accompanying article describes
an attempt to partially meet the criticism by utilizing, for
forecasting work on the Pacific coast, the charts published
each month in the MONTHLY WEATHER REVIEW. No working
meteorologist will fully agree with Doctor Schuster’s opinion
exprest above; yet the need of further study of the data
now accumulated is evident and the limitations of our present
methods manifest. And yet, has not too much been expected
in the matter of forecasts. I f not at the present time, cer-
tainly in the past, results have been expected entirely incom-
mensurate with the facts and data furnished. Nor is there any
present method of verification which does or can do full justice
to the forecaster.
I n recent years the recognition of the part played by the
larger pressure areas, the so-called permanent and subperma-
nent cont.inenta1 and oceanic areas, has given the forecaster a
possible means for undertaking seasonal forecasts with some
prospect of succecls. The importance of extending the area
of reports is now more than .ever recognized. With the ex-
ception of the exploration of the upper air, the study of sea-
sonal displacements of the areas of sea-level pressure offers
the most promising field for helpful work in forecasting.
Over the Pacific Ocean, plainly, not less but more observa-
tions are needed. Absence of reports now handicaps the fore-
casters on the Asiatic as well as on the American side of the
Pacific. It is conceivable that with a close working coopera-
tion be tween the Japanese, Indian, Chinese, and Philippine
weather services and those of Mexico and the United States,
including Alaska and British Columbia, aided by the receipt
of wireless weather measages from vessels at sea, the fore-
casting officials of these services would bein a position to un-
dertake general forecasts for a period of a week or longer,
eventually determining seasonal forecasts. And it may not
be amiss to call attention to the excellent work done in fore-
casting on the Pacific coast, and to say that, valuable as the
daily forecasts have been, the same degree of efficiency for
h L . 1908 . MONTHLY WEB- BEVIEW . 99
longer period forecaste would be of much greater value . The
importance of seasonal forecasts for the Asiatic districts is
widely recognized . It is not so generally knovn that there
are well marked dry and wet periods in southern California
and Arizona; and that some knowledge of the likelihood of
abnormal conditions. especially dry winters. could be directly
utllized by farmers and stockmen . In this paper a description is given of a device. a glass map
tray. which meeta to some degree the requirements alluded to
above: 1 . Extending the area of reports . 2 . Permitting com-
parison with typical conditions . 3 . Permitting the study of
of seasonal displacements of sea-level pressures . While the
immediate problem was to bridge the Pacific Ocean. it ap-
pears that the method can be used advantageously in bridging
the Atlantic . In the illustration (fig . 1. not reproduced) there
is shown a series of glass maps for the month of January.
1908. covering the United States and extending eastward
over the Atlantic and over Europe . In all an area of many
caster . A basic fact in meteorology. namely the general drift
of the lower air from west to east is utilized in this device.
by either moving each map a proper distance eastward every
twenty-four hours or by moving a skeleton map of the United
States an equivalent distance westward . The positions of high
and low areas during and subsequent to their passage acrosm
the continent are thus shown . Forecasters on the Atlantic
coast are enabled to follow the pressure areas long after these
have past beyond the limita of land report a Under the second
requirement. comparison with typical conditions. fig . 1. shows
a typical wet month. January. 1896 . The sea-level isobars.
isotherms. and resultant winds for any month. as charted in
the MONTHLY WEATHER REVIEW. are thus instantly available for
study . The normal pressure on the sea-level plane taken from
the Report on Barometry can serve as a base for pressure
departures; altho in the work at San Francisco where morning
and evening forecasts are issued. it is necessary to use cor-
rected normals . An enlarged portion of the frame is shown ..
million square miles is thus spriad before the eye of the fore- on fig . 2 (not reproduced) .
TABZE 1-Sun Francisco mi
SeaSon . I July .
184%'50 ................................................ 18W51 ................................................ 1851.52 ................................................. 1852-'58 ................................................. 1853-54 ................................................. 18-55 ................................................. 1855-'66 ................................................. 1856-'67 ................................................. 1857-'68 ............................................... 1858-'59 ................................................ 18j9-'60 ................................................. 186&'61 ................................................ 1961-'62 ................................................. 1882'63 ........................................... , .. 1863-'64 ............................................... 186445 .............................................. 1865-'66 ............................................... 1866-'67 ................................................ 186i-'68 ................................................ 1868-'69 ............................................... 1e69-'70 ................................................. 1870-'71 ................................................. 1871-'R ................................................. 1872'73 ................................................. 1873-'74 ................................................. 1874-'75 ............................................... 187S-'76 ................................................. 1876'77 ............................................... 1877-'78 ................................................ 1878-'79 ................................................. 1879-'80 ................................................. 1880-'81 .............................................. 18rll-'82 ................................................. l882-'s8 ................................................ 1883-.84 ............................................... l884-'85 ................................................. 1RYS-'86 ............................................... 18%-.87 ................................................. 1887-'88 ................................................. I888-'89 ................................................. 1.190 ................................................. 1@30-'91 ................................................ 1891-'92 ................................................. l892-'93 ................................................ 189w94 .............................................. 18~-'95 ................................................. 1895-'96 ............................................... 18W97 ................................................ 1897-'% ................................................. 189R-'99 ................................................. 1899-'1900 ............................................... 190Ck'Ol ............................................... 1901-'02 ................................................. 1902-'03 ................................................. 1905-'04 ................................................. 1904-'06 ................................................. 1905-'06 ................................................ 1906-'07 ................................................
Average (58 years! .............................
0 . 00 0.00 0.06
0 . 00 0.00 0.00 0.00 0 . O?
0 . w 0 . oj
0 . 00 0.21 0.00 0.00 0.00
0.00
0 . w 0.00
0.00
0.00
0.00
0.00 0.00
0.01 0.01 0 . o(I
0 . w 0.01 0.09
0.01 0.01
0.00 0.00 0.00 T . 0.06
T . 0.01 0.01 0.02 0.10 0.00 0 . 02
T . 0.01 0.04
T . 0.00 0.00 T . T . T . 0.00 0.02
0.00
0 . a3
0.02
n . 00
o . as
Aug .
0.00 0.00 0.02
0.00
0.04 0.01 0.00 0.00 0.05 0.16 0.02 0.00 0.00
0.00 0.00
0.00
0.00 0.00 0.00 0.00 0 . 00 0.02
0 . os 0.00
0.00
0 . u1 0.00
T . 0 . O?
0.00 0.00 0.00 0.00
0.04 T . T . 0.01 0.01
T . 0.00 0.02
0.00 0.00 0.00 0.09
T .
1'.
T . T . T . T . 0.06 T . 0.11
0 . 02
-
o . a i
o . no
am
fr .
h P t
.
0 . 00 0. 3.. 1.03 0.00
0.46 0.15 0.00
0.0;
0.00
0.00
0 . *3
0 . 00 0.02
0.00
0.01 0.24 0.11 0.04
0.00 0 . l? 0 . m 0.00 0.04 0.00 0 . 02 0.00 0 . 38 0.00 0.55 T . 0.00 0.35
0 . !&i
0.42
0.33 0.11 0.01 0.29 0.98
T . 0.31 0.77 0.02 0.21 1.05 0.77 0.12 0.10 1.06 0.00 0.46 0.78
7 . T . 6.07 T . 0.18
0.30
am
all. dig, eeasanal. and annual. 184%1907 .
Oct.
.
3.14 0.00 0 . ?I 0 . so 0 . I? 2.43
0.00 0.45 0.98 2 . 74 0.05 0.91 0.00 0.52 0.00 0.13 0.26 0.00
0.15 1 . ?9 0.00 0.07 0.11 0.83 2.69 0.24 3.36
1.27 0.78 0 . a5 0.54 2.66 1.48 2.65 0.72 1.48 1'.
0 . 14
0.00 0.01
0.16 1.73 0.11 1.55 1.70
0.86
3.92 1.48 0 . G4
1.70 0.17 2.37
1 . 0.03
1.02
am
n . rfi
7 . as
1 . r i
Nov . -
8.66 0.92 2.12 5.31 2 . ?S
0.34
0.67 ? . i 9 3.01 0.69 7.28 0.5s 4.10 0.15 2.55 6.68 4.19 3.35 3.41
1 . I8 1.19 0.43
? . 81 2.79 1.16 6.55 7 . '27 0.55
1.57 0.57 4.03 0 . %3
1.94 4.18 1 .BO
0 . ?6
10.00 0 . BL
0.99
8.99 2.90
0.00 0.56
3.91 4.18 0.88 1.78 4.56 1.05 0 . 46 3.79 3.91 3.18 1.98 4 . 35 1.07 0.92 1.69
2.66
The third requirement. that of quick recognition of sea-
sonal pressure displacement. is met in the following manner:
Records of many years are assembled . Rainfall tables em-
bracing a period of nearly sixty years are prepared for at least
three stations in California (e . g., Table 1. San Francisco
rainfall. 1849-1907) . While these printed tables give only
the intensity or amount of precipitation. auxiliary tables give
the frequency. or number of rainy asp. Thus the forecaster
Dec . -
6.20 1 . 05 7.10 1 3 .a ? . 32 0.87 5.76 3 . i 5 4.14 6.14 1.57 6 . 16 9.54 2 . %5
1 . RO
8.91 0.58
16.16 10.69 4.34 4.31 3.88 1 4 .1 5 . $5
9.7? 0 . 33 4.15
0.00 2.6G 0.58 4.46 I'z . 33
8.85 2 . 01 0.92 7.68 4.99 2.07 3 . .u 5.80 I3 . 81 3 . 25 5.62 5.08 2 6 9.01 1.43 4.34 1.22 1.62 2 6 5 1.37 0.90 2.82
1.68 1.59 2.05 6.90
4.68
-
Jan . -
8.34 0.72
0.66 3.92 3.88 3.67 9 . M
2.45 4.36 1.28
1.64 2.47
3 . m 1.83 5.14 10.85
5.16
9 . w 6.35 3.89 3.07 4.00 1.58 5.66 R . 01 7.5s 4.3? 11.97 3.52 2.23 8.69 1.68 1.92 3.94 2.53 7.42 1.90 6.81
1.38 9.61 0.98 2.12 8.05
5.99 6.99 8.14 2.26 1.12 3.67 4.11 5.79 1.23 3.78 1.05 4.04 3.90 4.41
4.72
a4 . s0
-
Feb . -
1.77 0.54
0.14 1.42 8 . M
4.77 0 . .w 8.59 1.83 6.39
1 .IN 3.72 7.53 3.19
0.00 1.31 2 . I?
6.13 3.90 4.78 3.76 6.90 8.94 2 . 21
1.18
4.90 1.87 4.65 2.96 1.04 6.65 0.30 0.24 9.24
0.94 0.72 5.16 7.26 2.90 2.75 2.69 2.31 0.28 4.41 2.13 0.10
0.64 6.03 7.27 1.76 6.89 2.70 4.30 8.02
3.54
7 . m
o . 32 4 .sa
i a . sa
-
Bar&
.
4.53 1.94 6.68 4.86 3.51 4.64 1.60 1 . e2 5.55 3.02 3.90 4.08 2.20 2.06 1.52 0.74 3.04 1.58 6.30 3 . 14 2.00 1.31 1.59 0.79 3.36 1.30 6.49 1.08 4.66
2.011 0.90 3.45 3.01 8.24 1.01 2.07 0.84 3.60
7.78 4.73 1.96 2.85 4.08 0.60 1.89 2.85 4.56
0.81 7.61 1.91 0.80 2.65 6 . 23 6.01 3.15 5 . 02
8.4
3.87
IJ . m
April .
0.46 1 . 3?
0.26
5.37 1 . 12 5.00 2.94
0.00 1.55 0.27 3.14 0.51 0.73 1.61 1.57 0.94 0.12 2.36 2.31 2.19 1.53 1.89 0.81 0.43 0.90 0.10 1 . m 0.46 1.06 1.89
10 .OB 2.00 1.22 1.51
G . 33 3.17 5.38 2.30 0.11 0.96 1.18 2.44 1.39
1.0.7
0.50 1.24 5.16 0.27 0.19 0.62 1.08 1.64 0.98 0.56 1.29 1.33 0.92 0.11
1.74
May .
0 . OI 0.6i 0.35 0 .3 0.05
1 . %
0.7$ 0.05 0.34 1 ..%
2.86
1.00 0.74 0.22 0.78 0.63 1 . M
0.03
0 . ?3 0.18 0.00 0.60
0.24
0.18 2.55 1.14 0.21 0.21 1.6s 0.2.3
0.04
0.87 0.06 0.58 2.17 1.07 1.25
1.86 0.15 1.31 0.60 0.72 0.61 1.44 0 . e6 0.32 0.69 1.05 T . 0.30 2.06 2.76
0.04
0.75
aoa
o . oa o . m
0.24
o . is
June.
0.00 0 . 02 0.00
0.00 0.08
0.00 0.03 0.12 0.05 0.00 0.09 0 . OR 0 . a5 0.00 0.00
0.00 0.04 0.00 0.23
0.00 0.01 0.04 0.02 0.14 1.W 0.04 0.01 0.01 0.05 0.00 0.69 0.04 0.01
0.19 0.01 0.07 0.27 0.03 0.10 0.11
T .
0 . (16
0.00 0.00 0.22
u . 19 0.01
0.05
T . T . T . 0.00 0.56 1 . 28
0.16
a oa
a . 67
o . a3
.r .
Jeaaonal
33.1(
IS.&
35.z ?3 . 8i z3.7t 21.w 19.91 21 .SI 22 . n 22.25 I9 . n
13.74 10 . oe 24.73 22.91 34.92 38.84 21.35 19.31 14.11 30.78 15.66 24.73 20.56
31.19 . 11.04 36.18 24.44 26.66 29.8B 16.14 20.12 32.38 18.10 31.33 19.04
23 . e6 45.85 17.58
21.75 18.47 25.70 21.25 23.43 9.58 16.87 I8 . 47 21.17 18.98 18.28 20.59
23.45
26 . li
n . 98
7 .1
ro . a1
i n . 74
ia 53
20 . 42
YeSr .
1850 1851 1852 1653 1- 1855 1856 1857
1858 1859
1 S O 111
1962 1863 1864 1865 1866 1867 1868
1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 187!l 1 1 0 1881 1882 1883 18.94
1885 1886 I S 7 1888 1889 1890 1891 1892 1893 1894 1895
18% 1897 1898 1899 1900 1901
1902
1903 1904 1 W 1906 1907
......
An- nual . -
17.40 15.60 27.29 21.17 22.45 26.39 2231
9 .4 6 21.39 21.18 25.52 38.63 15.10 21.64 14.06 36.28 50.64
8 0. 17 22.59 16.24 27.53 2 2 .4 18.66 22.62 n.65
23.M 11.93 .33. ari 30.76 30.07 23.73 18.67 15.48
50.a 23 . 18
m.02 19.04
28.03
36.94 25.43 21.11 22.08
17.91 24.82 17.18 28.25 16.40
8.51 a23 15.33 19.75 19 . 18 18.33
24 . 12 16.24
a0 .96
m . M
, ......
2217
has a ready reference table of wet and dry months . He has
also that which is of more value. viz. abnormal periods. or
months when there was little rain and months of excessive
rain during winter . He can also refer quickly to months of
excessive rain during the normally dry period . Confining
the discussion for the present to a single month (e . g., January)
it is seen that for the central portion of California there have
been seven abnormally dry periods in the heart of the wet
100 APRIL, 1908
the winter months. It is an interesting fact, but whether a
coincident or not we do not say, that a drought period oc-
curred in 1863-64, or at a time corresponding to the Briickner.
35-year period. Again, after nearly the same interval we have
another drought period 1897-98. Crop yields, however, ara
influenced by many other factors than that of rainfall.
I n California there is a growth of Sequoia, both senipemyirens
and gigmitea, where rings of annual growth may be traced for
periods extending in some cases to nearly three thousand
years. I n these rings we have to some degree an integrated
history of the seasons and possibly the periods of extreme
drought and excessive rainfall could be determined. The tree,
however, is not a ready witness and the study, when made,
must be entrusted to careful and competent hands.
I n communicating the preceding paper, which is printed by
special order of the Chief of Bureau, Prof. Alexander G. McAdie, says :
“As the paper touches upon the possibility of seasonal fore-
casts on the Pacific coast, or rather upon forecasts for periods
of ten days or longer, I would appreciate a personal reading
by the Chief of the Weather Bureau and an exprerrsion of
opinion from him, as well as any comment or criticism which
my colleagues in the Forecast Division may care to make.”
Accordingly the following remarks are appended:
While it is f a r from true that conditions of pressure and
temperature presented by a map of one day will represent con-
ditions that will exist 16O, more or less, to the eastward on
the following day, yet the tracings on glass are, in a measure,
useful. The paper, as a whole is interesting, and its publica-
tion is recommended.-”. B. Q.
As yet I am very skeptical as to the possibility of extending
our period of forecasting effectively, chiefly, as stated by
Professor Henry, because of the want of persistence of any
well-defined type of pressure distribution, and because of the
comparative uncertainty as to the direction of movement of
areas of high and low pressure, even for twenty-four hours in
advance. However, I think Professor McAdie’s paper a very
intoresting one, and of value that will fully justify its publica-
tion.-H. C. F.
The relation of the pressure distribution for several months
to the weather for the corresponding period has been a fruit-
full subject of discussion for some years. The work by Hilde-
brandsson, de Bort, and Hann, in Europe, is well known.
These scientists have shown that the weather of northwestern
Europe is related in a general way to the pressure distribu-
tion over the Azores and in the vicinity of Iceland, respec-
tively.
The author of the paper now under consideration calls
attention to the apparent connection between the monthly
pressure distribution on the Pacific coast and the rainfall;
also to a device for projecting the current distribution of pres-
sure forward over 15’ of longitude, and discusses the utility
of this device in forecasting.
That the weather experienced from day to day depends
almost wholly upon the pressure distribution goes without
saying, but when an attempt is made to apply the knowledge
thus far gained as to the influence of the so-called permanent
and subpermanent continental and oceanic areas on the
weather of both adjacent and somewhat remote regions, the
same difficulties are met that are encountered in the twenty-
four hour forecasts, viz, the inability of the forecaster to fis,
with any degree of certainty, the time that a condition once
establisht will persist. I speak now more especially of the
so-called continental areas of high pressure and forecasting
in the United States. The movement of highs and lows in
this country, especially in the cold season, is far too rapid to
permit of the formation of areas of either high pressure or low
pressure that can properly be called even subpermanent.
season; namely, in 1851, 1852, 1889, 1891, 1898, 19Q2, and
1904. For some of these we have the monthly oharts of
isobars available, and by making a composite we can formu-
late the following general laws for forecasting:
A When the continental high overlies Oregon, Idaho,
Utah and Nevada, the general drift of the surface air is from
the north or northeast; and such a circulation favors fair
weather, with little precipitation. Individual highs are likely
to move slowls eastward. Individual lows are restricted to
northern counties, and pass as a rule eastward without ex-
tending southward.
From the nine abnormally wet seasons, viz., 1850,1856,1862,
1866, 1875,1877, 1881, 1890, and 1896, we learn, using such
charts of the MONTHLY WEATHER REVIEW as are available, that-
B. When the north Pacific low area extends well southward
along the Oregon coast and the continental high overlies
Assiniboia and Montana, the general drift of the surface air in
California is from the south or southeast. Conditions favor
unsettled weather, with frequent and heavy rains west of the
Sierra and heavy snowfall in the Sierra. Individual highs
appear with little warning north and east of the Kootenai, and
move as a rule slowly south. Individual lows appearing over
Vancouver Island and the north coast of Washington, deepen
and also extend southward, the rain area reaching northern
California in twelve hours, the central coast in twenty-four
hours, and the coast south of Point Conception in thirty-six
hours.
. Combining A and B we can estimate the relative change in
pressure and air movement for a given increase of precipita-
tion. I f records of duration of cloudiness and rain were avail-
able, some relation might be found between the direction and
strength of surface winds, and duration of rain.
Taking up the abnormal periods for February, we find some
notably dry months, e. g., 18G4, when the entire month was
without rain; 1850, 1852, 1856, 1886, each with four rainy
days; 1875 and 1883, with three rainy days, and 1889, with
but two rainy days. Other dry Februaries were 1885, 1896,
and 1900. We have no records for 1864; but we feel able
because of these studies to chart the probable pressure distri-
bution for that month; namely, unusually high pressure over
Idaho, Oregon, and Washington, with surface winds from the
north and east.
With regard to wet periods, we find that the following
Februaries were abnormally wet, 1854, 1857, 1869, 1862,
1867,1878,1887,1891,1902.
If we consider the number of rainy days, rather than the
amount of rain, we have the following as abnormally wet Feb-
ruaries: 1854 (16 days), 1857 (15 days), 1859 (18 days), 1872
(17 days), 1873 (17 days), 1878 (19 days), 1887 (16 days), 1891
(19 days), 1897 (17 days), 1902 (19 days), 1904 (16 days).
Composites of these confirm the law given above under B.
As illustrative of the difference in the amount of rain falling
in a dry and a wet month, we give the isohyets for California
for January, 1902-a dry winter month; and for February,
1902-a wet winter month (see Charts IS and X).
For the dry month there was an estimated deficiency, de-
termined from normals for 194 stations, of 81 millimeters (3.17
inches) or approximately 35,755 million tons of water.
For the wet month, the excess of precipitation was 131 milli-
meters (5.17 inches) or 58,320 million tons of water.
California offers exceptionally good opportunities for study-
ing seasonal variations and climatic abwrmalities. Supple-
menting the comparatively brief record of nearly sixty years,
we have meager records of extremely dry periods noted by
the failure of crops during the time when the Mission Fathers
controlled agricultural and stock interests. There was, for
example, a failure of the crops at all of the missions in 1829,
* * * * * * *
we naturally infer that there was a period of drought in
BPBIL, 1908.
Even the so-called Plateau high, ” an area of high pressure
that occasionally lodges over the Great Basin in winter, does
not persist on the average over two or three days. There are
times, of course, when it persists for a longer period, but on
these occasions it is believed that the endurance beyond the
average period is due to the inflow of cold air from the north-
west. In other words, the Great Basin, owing to the topo-
graphic surroundings, actually serves as a basin or reservoir in
which part of the cold air which has a slow eastward or south-
ward motion is entrapped.
One of the reasons for entertaining this view is the fact that
offshoots from the “Plateau high ” are frequently discharged
to the eastward or southeastward. After the discharge of an
offshoot the parent high soon disintegrates. It is also be-
lieved that owing to local radiation, and the drainage of cold
air into the valleys occupied by Weather Bureau stations, the
sea-level pressures for the Plateau region, are at times greatly
affected by these local surface temperature falls.
An area of high pressure firnily lodged over the Great Basin
is a most important asset to the forecaster, not only on the
Pacific slope, but eastward from the Rocky Mountains includ-
ing the Northern and Central States.
As I said before, according to my experience, there is no
distinct pressure formation in this country that approaches a
condition of subpermanency. There are times when the highs
and lows follow each other in nearly the same path. When
this condition prevails it is customary to say that a certain
type prevails. The same phenomenon has been observed in
Europe: See Nils Ekholm in the January, 1907, Meteorolo-
gische Zeitschrift, G c Uber die Unperiodische Luftdrucksch-
wankungen und einige damit zusammenhiingende Erschei-
nungen.” (On the nonperiodic pressure variations and some
phenomena in connection therewith.)
. It seems to me that it would be profitable to study the daily
weather maps in periods of less than a month since the latter
period is too apt to include the records of more than one type.
It might be possible to do this for the Pacific coast where the
atmospherio movements are less complicated than in eastern
districts.
As an illustration of what one would meet in attempting to
correlate monthly mean pressures and weather conditions, I
submit, herewith, copies of the monthly mean pressures for
March, for the years 1902 to 1906, together with the paths of
highs and lows on the Pacific coast and over the Plateau
region. But one of these months (March, 1904) shows a steadiness in the movement of lows that would be useful to
the forecaster. (Charts SI to SV.) Aside from the forecasting point of view considerable
interest attaches to this subject on account of its bearing upon
a rational explanation of climate. In this connection see
Bulletin Q, under ‘‘ Seasonal variations of the weather.”
In regard to Professor McAdie’s second proposition, I would
say that we are never certain that the pressure distribution
and the weather conditions existing at any moment of time
will match the actual conditions to the eastward in the nest
twenty-four hours.-A. J. H.
101
UAN W E PROTEUT AGAINST TOFWADOEST
A well-constructed conductor is a fairly reliable protection
against destruction by lightning, but one must be inside the
protected building, as there is no assurance of safety on the
outside. A dwelling may be so constructed as to pass unin-
jured thru a hurricane, tornado, earthquake, fiood, or fire, tho
it is rare that suah are built, and that which is safe against
one kind of visitation may not be so for another.
The following correspondence shows one phase of the ques-
tion of protection. We should like to have some one compile
15-3
enough data to give us a fairly correct idea as to whether it is
best to be frightened at every storm cloud and run to the
shelter of the c c tornado cave,” or whether we may not as well
be brave and calmly await the dread visitation, since it is most
likely to pass us by. As the result of his extensive studies
Lieut. John P. Finley maintained that the best we oan’ do is
to watch the distant tornado, and if it seems to approach us
then move away toward the left; so far as we have learned,
this still continues to be the best rule.
(1) LETTER FROM A CORRESPONDENT TO THE CHIEF OF BUREAU.
I am going t o establish in this city a system which will give us warn- ing of the approach of tornadoes, which is as follows:
We will run a pole line around our city at a distance of four miles, which is connected to an alarm in the city by wires, using the very best
of wire and puttlng up the line in the most substantial manner. There will he two wires on this circuit around the town. We will place instru- ments a quarter of a mile apart on this line, to be adjusted to short- circuit the wires by making an electric contact, should a change in the atmosphere pressure (of as much as three-tenths of an inch) take plaae within flve minutes.
The magnetic apparatus for giving an alarm in the city is arranged 80
that if one or both the wires are broken it will cause an alarm to be given, or should the wires touch one another by being twisted together
it would give an alarm; also should the instrument short-circuit the wires by the sudden change in the air pressure, we would receive
alarm. We will have notice in advance of the tornado by the time it would take i t to travel from the instrument or pole line to the city, and
as the line is all around the city, at a distance of four miles, it would be uuable to reach the city from any direction without giving us an alarm.
I have kept in touch with the great work that your Bureau is doing under your able management, and I earnestly hope that you and your good workers may live to quite an old age, as you have done much to
overcome ignorance and superstition in regard to the many fake idem of the people in regard to forecasting the weather.
REPLY TO THE ABOVE LETTEB BY THE EDITOR.
You propose to surround your city by a double line of telegraph wire’
inclosing an area about 8 miles in diameter or 25 miles in circumference.
A t every quarter mile of this circumference (100 stations in all) you will place an apparatus that will automatically shoi-kafrcuit the line whenever
the atmospheric pressure rises or falls at the rate of three-tenths of M
inch in flve minutes, or faster. The wires will also be short-circuited,
or an alarm given, i f either wire is broken or if the wires touch each other. You think that thls system of alarms will protect the city from
the unexpected arrival of a tornado. The statistics of tornado frequency show that a region of 1 square
mile any where in Missouri ie not very apt to experience a tornado, the probability being one-sixteenth of one per cent per century, so that a region 4 mlles square would have a probability of 1 per cent per century; that is to say, i t will presumably have a tornado once in ten thousand years. It is therefore probable that your system of wires and apparatus would have t o be kept In working order many years before it could be of use; and unless i t be kept in perfect repair, at great expense, i t will
be out of order and useless when the tornado comes. For these reasons many schemes analogous to yours which have been proposed in the last
forty years have been abandoned as impracticable. If you have any statistics showing that tornadoes are specially frequent in the neighbor-
hood of . . . . . . . . . . . . . . I should he glad to receive them and revise this calculation.
Ynu state that you are going to establish your system around . . . . . . . .: does this mean that you are going to do it at the expense of the city, or as a private enterprise?
(3) SECOND LETTER FROM A CORRESPONDENT.
In reply to your letter, I will say that I am unable to give you any additional information in regard t o the frequenoy of tornadoes in this
State, for the following reasons:
1. That most newspapers, as well as the public in general, call tornadoes
cyclones.”
2. I have only been able to visit a few of the damaging storms to mcertain what they were.
3. That I believe that only a small per cent of the tornadoes reach the
earth’s surface, and i t seems to me that it would be dimcult on this account t o ascertain or even attempt to approximate the number of tor- nadoes in any locality. I quite agree with you that i t is only a small per cent of tornadoes
that do damage in this State, but there are so many dark, dangerous, and threatening-looking clouds that we will be uneasy, and this will
cause us to always keep our alarm system in the best of condition for fear that they are tornadoes. I have only this one life to live, and being healthy and enjoying it
there is no expense I would not undergo to protect it, 88 I prefer living
to any other state or condition I can imagine. We will not install this