MAY, 1981. MONTHLY WEATHER BEVIEW.
A NEW CORRECTION-SCALE FOR MERCURIAL BAROMETERS.
By S. P. FERQU~SON. sc/. 5*&
Weather Bureau, Wnshington, D. C., June ZO, 1921.1
289
SYNOPSIS.
Corrections. for inst.rumenta1 errors, the temperature of the mercurial column and for latitude, must be applied before reading8 of mercurial barometers can be used to represent true at.moephenc pressures. Ob-
viously. each operation is a possible source of error, and a direct met,hod of determining correct values is highly deGrable. In 1911, Col. E. Gold, of the British Meteorological Office, suggested a scale to be used with the attached thermqmeter, by means of which the corrections referred to can be obtained in one reading of the ther- mometer. This scale. however, c?n be used for but one pressure; in the present paper the author describes a lnodlfied scale from whlch the same corrections can be obtained for any pmsure. The final corrected pressure is obtained in five or less operations instead of six required by ot.her methods, the correct.ions can he read more accurately, and in addition to the time saved there iu a saving of one column of entry in the record-form.
The mercurial barometer, one of the most useful of
meteorological instruments, is one of the most incon- venient for ordinary use for the reason that, its indics- tiom or readings do not usually represent true atmos- pheric pressures until they have been corrected for tem- perature, latitude, and instrumental errors such as those of the scale, the “attached” thermometer, and capil- larity. The last three, for a fixed station, are practically constant, but require to be redetermined from time to time. Nine distinct operations must be performed in order to obtain the actual pressure at any place, and if this is to be corrected for height, three more are required; operation is a possible source of error and must be veri ed if the final values are to be depended upon. Mechanical methods of obtaining true atmospheric pressures have been proposed, the most satisfactory of which are the various forms of “syphon” barometers and devices for wei hing the mercunal column; but most
portable, possess defects as difficult to allow for as the errors of the sim ler patterns, and should be used only
known at the present time are those in which the pressure is determined by weighing and the various corrections are applied automatically and continuously. Of various methods that have been suggested for simplifying the process of determininu true pressures from readings of mercurial barometers, &at of combining dl constant corrections with that for temperature in a single table appears to be the most satisfactory. In some instances where the height is small the correction for hei ht may be included without serious error: also,
lower than its original location a “removal” correction may be included m order that the observations at the two positions may be comparable. This method of combining corrections has one serious defect-a new table must be repared for every barometer every time
tions become necessary; also, while three of the nine operations referred to are eliminated, the rocess is
ment itself. Methods more direct and more accurate than those referred to have always been desirable, but no important advance ap ears to have been accomplished until 1914,
when Col. E. Gold, of the British Meteorological Office, suggested a device whereby the several corrections already referred to may be combined in a single reading
barometers em E odylng such devices are not easily
K
b experienced o E servers. Exce tion must be made in t i? e instance of barographs, of w ich the most accurate
when t fi e barometer is moved to a new position higher or
the height of t 1 e mstrument is changed or new correc-
mdirect, and not automatically a function of t ?l e instru-
of a movable scale used with the “attached” thermom- eter.’ This device, to which the name “ideal scale” was given by the inventor, is illustrated schematically in
i
Figure 1, which is a copy of the original drawing, and can best be described by the f o l l o w quotation from Col. Gold’s paper:
As we do not use an ett,ached thermometer to give ut a temperat- which we wish to compare with other temperatures, but merely to enable UE to correct the reading of the barometer, it would beie con- siderable advantage to have $e thermometer so graduated that ita readings gave the correction dmct in millibars; in other words, ta
per my “Barometer Readings in Absolute Units and The& ’correct& m’a Rxction 3 published In the Quarterly Joumlof the Societv July, 1914. Thlspnper isvpnydlunble as n summnry of present-dav informntion re the measurement 01 pnasure and for its descriptions 01 ingenfous end origha&ds of wrracting obermtIons, eta.
1 Described before the Royal Meteorological Sodety Mng 20 1914 in a
rn MONTFILY WEATHER REVIEW. &Y, 1991
make a thermometer for which each male divieion carresponded with a correction of 0.1 mb. for a reading of 1,OOO mb. The divirriona would be about 10 per cent further apart than the divieions for degrees Fah-
would be especially advantageous for use at a f i x r station at a low level or for a barometer at sea, because, by an ap ropnat.e.adjustment of the scale by which +e thermometer is read, &e combined correo tim would be read off dmctly from the instrument.
For a fixed station the ecale would be set once for all at the appm priate latitude and altitude. It could then be clamped and the ob-
m e r would t h e n c e f o d emply take his barometer,reading and read off the necepsary correction from the ‘thermometer. The correction would need modification only i f the preasure differed considerably from 1,OOO mb.; usually, the correction, as read would be au5cient. * * In figure 1 the scale of height at the right-hand aide of the
diapam is supposed to be movable, relative to the latitude scale: whde the whole piece carrying the latitude scale and the ideal scale for the attached thermometer IS movable, relative to the thermometer. The thermometer itself k~ supposed to be fixed to the frame of thr barometer, on which is engraved the arrow shown at the right. To
B& for a reading. the height scale is moved until the ‘IO ” cpmes opposite the Iahtude of the place of observabon; the whole piece IS then moved until the number denoting the height of the barometer cistern M. S. L. in metrw comes opposite the arrow. The thermometer reading is then taken from the ideal seale, which gives the coirection for tem eratnre, gravity, and altitude combined for a preseure of 1,OOO mb. &is illus- trates the theory of the method. For practical application Mr. Whipple haa pointed out that simplification is introduced by fixing the latitude d e , relative to the thermometer, and making a single movable piece bearing the ideal scale and the height scale.
It was obvious, on reading the description thereof tnat modification would be necess to adapt the original
barometers whose height may be changed. The error resulting from the use of a single uniformly raduated scale through a range of ressure of SO mb. fihe range
large to be disregarded and which must be allowed for
b applying corrections or by the use of a mechanical djustment. As in the instance of the table of ‘‘totnl correction” already described, a separate scale woulcl be necessary whenever the height of the barometer is changed. After a few experiments with adjustable scales the tangential scale shown by figures 3 and 3 suggested itself and two e,Tperimental models constructed during November 1914 met all-re uirements very satisfac torily.
machine prevented further experimenting until recently, and the few scales produced have not been tried outside the laboratory; but i t is believed that the pattern described herein is not likely to require important modi. fication unless in tshe direction of greater simplicity. -4ssuming that the barometer in use is the well-known “standwd” of either Fortin or Kew pattern, and of good quality, the most important of the operations incidental to obtaining true pressures from its readings is that of
determining the correction for temperature. This is zero at a certain tem erature, which is 0’ C. for the Interna- tional and 28.6O%’. for &e English scale, at all pressures, and obviously its value a t higher or lower teniperatures
is directly roportionate to the length of the barometric column. $he construction of a correction-scale for any definite pressure, such as 1000 mb., has been described in the quotntion from Col. Gold’s aper. A scale for any
mb., is obtained in the following manner: First, the dis- tance, from the common zero, of an extreme value, such
as 5.0 mb., a t each of these ressures, corres ondin to
two parallel columns of dote (one for each pressure), cat
any suitable distance apart. By ruling lines connecting dote representing the same value we produce the tan-
gential scalc (B) shown in figures 2 and 3. The lines are
renheit. Such a thermometer is shown in the d’ (fig. 1). It
ideal scale to use throu h the 7 arge ranges of pressure experienced near the fi orth Atlantic coast or with
recorded a t Boston) is a f out 0.2 mb., a quantity too
Pressure of other work an 1 the lack of a suitable ruling
range of pressure, as, for esamp P e, that of 900 to 1000
the equivalent readin of the t R ermometer uselis mased on the blank to be ru 5 ed: next, this space is divided, by F/G 2, /NlU /BARS) SM,19M. F /G . 3, (INCHES)
&Y, 1921. MONTHLY WEATHER REVIm. m
International units.
Attached Ob8erved Total Station therm. reading. correction. pressure.
uniformly spaced, but since the values of corrections for low ressurea are smaller than those for higher pressures!
as the pressure decreases. Referring t.0 figure 2, t4he tem- terature equal to a correction of 5.0 mb. at a pressure of 1000 mb. is 30O.5 C.; that for a pressure of 900 mb. is 33O.8, or, otherwise expressed, the correction a t 900 mb. for a temperature of 30O.5 is 4.5 mb. instead of 5.0 mb., the value for 1000 mb. To obtain values for pressures between 900 and 1000 mb. vertical lines for any interval desired can be ruled on the scale itself, or preferably, as shown, on a transparent! index- late (C), which ca.n he
numbers of t8he vertical lines are always near the central index-line, where the readings are made. The index (C) slides on a guide (G) arallel to the scnle (B) and the
base (A), fitted and clamped to the outer metal tube of the barometer (T) by screws (S). To obtain corrections the central horizontal line on the index- late (C> is set
rend at the point re resenting the current pressure: as
correction for a pressure of 955 mb. is -2.82 nib. In figure 3 the correction for a temperature of 66O.7 and a pressure of 39.51 inches is - 0.1015 inch. Bllowance for constant errors or corrections already referred to is ac- complished as follows, the values even indicating the method only:
Correction for instrumental error. . . - - -. . . -. -. - -. . - - - -. +O. 15 mb. Mean correction applied to attached t,hermometer, -O.2’,
-. . -. - -. . +O. 35 mh. Correction for latitude. .. _.._.. . . .__. . .. .. . _. ..__._. . ._ -1. le mh. “Removal” correction .__._. -. -. . -. -. - -. . . - - -. - -. - -. - - - +O. 92 mb.
the B istance between the linea must be progressively larger
adjusted and possesses the furt K er advantage that the
thermometer tube. T K e various parts are secured to a
a t the top of the thermometric column an B the correction
shown in figure 2, wiere P the temperature is 17O.5, the
equal to __.___._. -. -. -. . -. . . . .
d t 8 .
Attaohed Obaervad Total Station therm. reading. correction. pressurs.
Total correction _.___.___. -. . . - - - - - - - -. -. -. - - - - - - +O. 20 mb.
This value, thereafter, is automatically added to all readings by moving the scnle (B) , until its zero is 0.20 mb. below the zero of the attached thermometer (0’ C.), and clamping it in position by means of the screws (E), (E). If, for esam le, the total constant correct,ion indicated is
to be adde a to the correction for temperature already obtained in the example shown for figure 3, the final value will be-2.63 instead of -2.52.
It is obvious that even if the ordinary “att.zclled” t;hermometer with a short scale is employed, these cor- rection-scales may be ruled direct to 0.1 mb. or 0.005
inch, and values to 0.01 mb. or 0.0005 inch estimated with considerable precision, an accuracy impossible with the ordin table of corrections determined for
Even in the extreme case of the use of a mountain barometer, where the correction-scale may be much compressed and the diagonal lines at high temperatures steep, the errors are smaller than those of the tables are vex escribed. Obviously, if there is no objection to increasing the bulk of the instrument, the tangential scale (B) can be made wider and ruled to fit a themom- eter with a long scale; but this is unnecessary except in the instance of a standard barometer where readings of extreme precision are desired. Ordinarily, the tangential
correction-scale need not occupy a preciably mom
eat hei hts,
where the ressure may be as low as 400 m f one siie of the scale-p P ate (B) may be ruled for pressures between
1050 and 600, and the other for pressures below 700 mb., and the clamp screws (E), (E), so placed that the scale
intervals of half 7 egrees and ten millibars or half inches.
space than does the usual “attached” t R ermometer.
When used with a barometer carried to
_______--
1860 w.55 -287 m08(
will be in the proper position with either side in me. The vertical lines on the transparent index-plate (C) will serve for both sideh of (B), and only m urn bo be
the other. Ordinanly, however, only one ruling will be needed for the average “station” barometer &t any heioht below 8,000 meters. 18ifferences between the customary method of reading the barometer and correcting the observations by a
table and that rovided by the new correction-scale are
To read the mercurial barometer and deterqnine the
“station” pressure by present method, recording dats in Form 1053 or form 1001-
(1) Read attached thermometer and enter reading in Form 1053.
(2) Adjust mercury in cistern. (3) Set vernier.
(4) Read vernier and enter reading in Form 1083.
(5) Ascertain total correction from table and enter in Form 1053. (6) Apply correction and enter “station” pressure in Form 1083.
In addition to the six operations indicated, four columns in Form 1001 are required for the observations find corrected data and there must be accessible a t all
times a special table of “total correction” for every barometer and every station and elevation; if any chan e of height or correction occurs a new table must ie pre ared.
$0 read barometer and determine “station” pressure by means of the tangential correction-scale, recording data in Form 1083 or Form 1001-
(1) Set index of correction-scale at current reading of attached thermometer, read and enter total correction 1 ~1
Forni 1053.
(2) Adjust mercury in cistern.
(3) Set vernier. (4) Read vernier and enter readin in Form 1083.
enter “station pressure” in Form 1083. Five operations as indicated, and three columns or spaces in the form are required. The correction-scale
forms a permanent part of the barometer, is instantly adjustable to any change of the corrections, and 1s ready for use a t any height. No separate table of cor-
rections necessary. Examples of entries in Form 1001, in English and International units when present method is employed:
numbered at the top for one scale and a t the I) ottom for
shown in the fo B owing comparison:
(5) Deduct total correction from o % served reading and
----
a70 xm -am mae
Iuternatlonal units.
Observed Total Station reading. Cor’n.1 pressurn.
_-____-
9ss.s -2.87 wit~
Engliah units.
Obwrved Total Statlon reading. m’n.1 pressure.
--
am -aim a8.m
!am MONTHLY WEATHER REVIEW. MAY, 1921
.me dassifioetion of o ations adopted may be ob-
tihe ape&ion of obtainin the correction from the scale;
-wed and &ere is larger robabilit of error in the
bon becomes necessary, than is the case with the new omection-scale. In the extremely rare instance of the need of the tem- par&ure d the attached thermometer, this can be ascer- tained in a few seconds by deducting the constants from
th9 total correction as recorded and looking up the equiva- lent, temperature. The precedin description and comparison may be
(1) The tangential correction-scale described is a means of determining the true pressure from observations of B
mercurial barometer wherever it may be used, without
jmted to for t b mason r at three items are included in
but mtual com arisen s % ows that much more time is
up of. 8 loose table, particular T y when K ouble interpola-
summadas B ouows:
(5) Errors in reading the attached thermometer, usu- ally very difEcult to detect, are entirely eliminated by the use of the scale, the index line of the transparent scale being set at the top of the thermometric column when a readin is to be made.
(6) Srrors in estimating values on the tangential scale are not so easily made nor are they so important as
errors in interpolating from the usual printed table of
corrections. (7) When the correction-scale is employed the column in Form 1001 for readin of the attached thermometer
where this temperature is desired are easily rovided for
correction. (8) As shown in the figures, the new correction-scale is constructed as a complete unit that may be secured to the outer tube of the barometer in the position usually
becomes unnecessary. !i! he extremely rare instances
by ascertaining the temperature correspon !i mg to any
srmvaa
F/G 4. /WCr/rmE FOR RULING TANG€ ffT/AL SCALKS.
the use of a separate table of corrections which may be lost or mislaid and is always an inconvenience. This scale forms a part of the barometer, as does the customary attached thermometer which it is intended to replace.
. (2) The scale can be adjusted to allow for constant &ors or corrections and the total correction obtained in one reading of the attached thermometer. Any change of correction is allowed for by a simple ad’ustment, and
(3) The number of operations required to obtain the “station” or actual pressure is smaller and the whole process sim ler than is the case when the customary
(4) Corrections for barometers of the ordinary-or sta- tion pattern can be read to 0.01 mb. or 0.0005 inch at intervals of 5 mb. or 0.05 inch; when barometers of pre- cision with widescale thermometers are- used smaller values are essil obtained. Equal precision can be se-
no adjustment is required for a change of h eight.
separate tab s e of correction is employed.
cured only by t z e use of special tables.
occupied by the attached thermometer and in the same manner. m e n the distance between the two screws (S), (S), (figs. 2 and 3) is known, the scalebase (A) can be bored accordingly and the correction-scale, complete, shipped to a distant station to be attached by the ob- server; i t will not usually be necessary to send barom- eters to a central point or a shop when scales are to be fitted or changed.
(9) The new correction-scale increases the cost of the barometer slightly, but it is believed that the advanta e
will more than outweigh the additional cost. The accuracy required in rulin the tangential scales is easily attained by the ordinary &riding engines usually found in shops and laboratones; but, because of the peculiar character of the ruling required, the paduation can be erformed much more easily and mth greater
and convenience of havin the means of correcting t % e
readings form an insepara % le part of the barometer itself
rapidity ?3 y means of the specid dividing machine shown
B ~Y , 1921. MONTHLY WEATHER REVIEW. !m
in Figure 4. essential1 of a
ivoted bar (B),
C), a plate (D), the blank to be ruled, and means which ma be either the micrometer (M), are mounted on a substantial cast-iron base (A). To ential scale, the limits of the s ace to be ruled
value, such as 5.0 mb. (determined by reference to the thermometer to be used), are marked on the blank which is placed on the plate (D) and so adjusted that when the bar (B) is in a central position the zero line of the scale to be ruled will be normal or perpendicular to the refer- ence lines &), &). During the process of ruling the blank is M y held by clamps (K), (R). .A wide range
or different positions of the clamps, also, the ate (D) may be secured in different positions on the ase,. (A), by means of screws (E), (E). In the figure, (0) indicates a partial1 ruled blank in position. The en raving mechanism T- C) can be adjusted to rule a
scse of unusual width b loosenin the screws (I), (I),
as may be desired; also, additional sockets, (S), are provided for the purpose of ruling scales with zeros near one end. If the scale is to be spaced by a micrometer (M), the edge of the bar, (B), is held ag.rinst a pin (P) in the slide of the micrometer, by a spring (not shown), so that when the screw is rotated the bar will move in the path indicated by dotted lines. For convenience in spacin , two rows of tapped holes (HI, (H), in the base (AT, afford means of securin the micrometer in any position by the clam s (K), (Kf A much P ess costly and almost equally accurate method 'of spacing is obtained by the use of a metal an le (R), on the ed es of which are cut ratchet racks of di d erent itches. 8 ith two sets of teeth of 2mm and 5mm pitck, res ectivel like1 to be nee ed can e obtained by clam ing t le rack in t K e proper position on the base (A). o avoid the slight error caused by spacing a di onal with a rack of appreciable thickness the perpendicu P ars of the teeth are
rounded slightly. All looseness or backlash is avoided by the use of a beveled knife ed e (N), which is kept in mesh with the rack by a stiff B at spring, and which is
easily-and quickly lifted and set in a new position by the knob (T). This ruling machine, including the rack but not includin the micrometer or the en aying mechanism, was bui k t from materials found in t r e instrument shop
of the Weather Bureau a t a cost smaller than the price uoted by manufacturers for a sin le tangential scale.
1 scale of either pattern described % erein can he ruled in about one hour, of which time the larger part must be devoted to adjusting the ruling machine.
ruling mec l anism
?
et rack (Rf These parts
and a the "- stance between the zero an a some extreme
is provided by additional tapped holes,
and moving the slotted ;g ase (U) % ackward or forward
7
correct spacing of eve
tf B E
sr/. 5 ; 6 2 8
RELATIONS BETWEEN WEATHER AND MENTAL AND PHYSICAL CONDITION OF MAN PRESENTED ON THE BASIS OF STATISTICAL RESEARCH.'
By ERNST BREZINA and WILRELM SCHMIDT.
[Translated and reviewed by W. W. REED, Weather Bureau, Atlanta, Ga.]
The behavior of the nervous s tern under different weather conditions is the chief su&ect of an investi a- tion having for test persons census clerks, school c f i 1-
1 SI&). Akd. Wfus., Wien. mat1iem.-naturw. Klassq Band 123, Abt. 3. 0ht.-Diz.
1914.
dren, and epile tics confined in hospital. Investigation
values and changes in value, also for general conditions such as distribution of pressure. These data were en- tered on cards daily for one year and opposite were placed comparative values characterizing erformance of mental work in office or school and cen 8 ition or be- havior in hospital. The following summary based on these data omits mention of those elements for which there appeared no plain relations. Relative to the influence of change in air pressure it appeared that a certain relation exists for li ht mental
pressure. Neither the number of epileptic patients affected nor the number of attacks shows any decided dependence on air pressure changes, and it is assumed that op osite results are to be referred to a simultane-
however, that there appeared lainly an effect of pres-
ing for the most part over 24 hours, unfavorable results accom anying the lar er amplitudes. It seems probable
in foehn sickness. Only the marked negative pressure departures proved bad for normal persons, while negative de artures generally were found so for e ileptics.
evident striking difference between the effect upon clerical force and that upon ersons affected with epi-
not well done at the time of high temperature or of marked tem erature departure, especially in the case of
to cold. For vapor pressure relations were found but little de- cided; there is to be noted, however, the generally favor- able conditions for office force and also for epileptics existin with the normally high vapor pressure of sum-
unfavorable. Much more decided results were obtained for relative humidity, and this element appears to have independent significance like that of air pressure and temperature. In winter the best clerical work waa done during high humidities (observations taken in the open), but this was due, of course, to their modi6cation to mean values by the heating of offices. Increase in the number of epileptic attacks with high humidity was very plainly shown in winter and was noticeable in summer. Low humidity was found to have bad effect. to the prevalent idea, a manifestation of influence y ozone was hardly to be recognized- and at
most only a slight effect was to he ascribed to tke wind, any influence other than mechanical is presumed ex- plained by other simultaneous conditions. The relation between meteorological elements that is founded on their connection with position of high and low pressure areas differed so reatly that it could not
observation and with region of rising pressure to the west manifested itself as most unfavorable for office
was made €or a P 1 meteorological elements, using current
work only; this probably proceeds best wit 5l uniform
ously c R anging condition of some kind. It is stated,
sure oscillations with periods o P 4 to 10 minutes extend-
that t R ese rapid osc' I f lations have essential signXcance
f!n the consideration of the thermal P actor there was
lepsy, which is summarized t R us: light mental work is
a duration o f two days; while epileptics appear sensitive
mer. % asimum values of ths element are, to be sure,
be accepted as ve serviceab B e. In general, however, the weather prevaiing '9 with fall region at the point of
interesting, especially so since weather effects are given