1887 .
] On Photometry of the Glow Lamp .
247 The absence of segmentation may probably be accounted for by the migratory habits of the embryo after development has commenced .
The fact that no sperm-morula is formed supports this view .
The evidence before us does not support the view that the ovum of Millepora formerly contained much yolk and has subsequently lost it .
I am inclined to believe that the Hydrocorallin\#174 ; belong to a separate stock of the Hydrozoa , which probably never possessed medusiform gonophores .
Millepora is not related to Hydractinia .
IV .
" On Photometry of the Glow Lamp .
" By Captain Abney , R.E. , F.R.S. , and Major-General Festing , R.E. , F.R.S. Received November 21 , 1887 .
In a paper which we read before the Royal Society ( ' Roy .
Soc. Proc. , ' No. 232 , 1884 ) it was shown when a carbon filament or a platinum wire in vacuo was gradually raised in temperature , that the different rays in the visible and invisible regions of the spectrum followed a law governing their intensity .
In the dark region of the spectrum ( below the red ) if the abscissae to a curve represented watts ( current X potential ) , and the ordinates the intensity of the ray under consideration , the curve so formed was hyperbolic , approaching more nearly to the parabolic form as the red was approached .
In the visible spectrum the parabolic curve was actually reached , the vertices of the parabolas moving along the axis of abscissas ; the shift being greater the more refrangible the rays under consideration .
This implied that until a certain number of watts had been expended the ray was absent .
Further , we had shown in the ' Philosophical Magazine ' for September , 1883 , that when measured by a thermopile , total radiation oc ( watts \#151 ; constant ) .
In the visible radiation of an incandescent filament in a glow lamp we are only dealing , however , with a small portion of the radiation , and therefore could not expect it to follow such a simple law as that which governs total radiation .
It appeared probable , however , that as the intensity of any individual ray in this part of the spectrum increased parabolieally , the sum of all the visible rays ought also to follow very closely the same form of curve , the vertex of such parabola lying at some point in the axis of absciss\#174 ; between the vertices of the parabolas of the extreme visible rays .
It likewise appeared probable that when the rays of extreme refrangibility were absent or in defect , as is the case when the filament is red hot , the parabola would fail to represent the intensity of visible radiation .
In the communication we have already referred to one example of 248 Captain Abney and Major-General Festing .
[ Dec. 8 , the applicability of the parabolic formula was given , for white light , but by itself it was hardly conclusive .
We , therefore , conducted a series of expeiuments to ascertain if our anticipations were correct .
An incandescence lamp was selected as a standard lamp , through which a fixed current was maintained .
This we used instead of a standard candle or other variable light .
We then selected a second similar lamp , of which to measure the light when currents of various strengths were passed through it .
The shadow and grease-spot methods were both experimented with , the former being perhaps the most exact .
Whichever method is , however , employed , it was inexpedient to move either lamp towards or from the source , or to vary the distance of the source from the lamp , as the carbon filaments show more or less illuminating surface to the screen according as they are close or distant from it .
It therefore became necessary to adopt some other plan for altering the intensity of the light falling on the source from the comparison lamp .
In the Rumford ( shadow ) method , fig. 1 will give the general idea of the arrangements .
Fig. 1 .
Ihe shadows cast by the rod D from the two sources of light , Ij1 and Lu , were made just to touch each other , on the white screen SS , 1887 .
] On Photometry of the Glow Lamp .
and to fall within a rectangle cut out of black paper , which deadened the light on the rest of the screen .
Each lamp ( Lq and Lu ) was in connexion with an ampere-meter and volt-meter ( Ax , An and Vx , Un ) .
In front of Iq ( the comparison lamp ) was placed an electromotor which caused a pair of sectors of variable aperture to rotate between it and the screen .
Evidently two methods are open to equalise the illumination of the screen from each source:\#151 ; 1st .
Cutting off more or less light from L1 .
2nd .
Varying the current in Ln by means of the variable resistance in the circuit .
The first plan necessitates the opening and closing of the sectors whilst rotating , and the second the alteration of the resistance , \amp ; c. , at will .
Whichever method was adopted the lamp Iq was brought to a bright yellow glow , and the lamp Ln had a current passed through it which , when the minimum resistance was in circuit in R , produced a brilliant white light .
Such intense heat the filament would not be able to stand for any considerable time .
When measurements were to be taken by the first plan the instrument shown in the annexed diagram , fig. 2 , was employed .
A pair of Fio .
2 .
9 1 \m sectors , S ( each of 90 ' ) , are mounted on a horizontal axis , a similar pair , S1 , are carried on a short sleeve , to which are attached two horizontal pins , passing through holes in the flange D of another 250 Captain Abney and Major-General Fes ting .
[ Dec. 8 , sleeve C. A stud in this last engages with a screw thread of long pitch cut on the axis .
A horizontal movement of C thus causes it , as well as the sectors S1 ?
to rotate with reference to the axis and the other sectors ( S ) , and therefore alters the aperture .
This movement is given by means of a vertical lever engaging with the groove K on C , and which is actuated by the screw B. The aperture can thus be varied between 0 ' and 180 ' , whether the instrument be in motion or at rest .
The instrument has been described at length , as it varies in some particulars from previous ones , having been made under our own supervision .
The edge of S is graduated into degrees so that the amount of aperture is at once known .
A certain current is passed through Ln , which is noted , and the sectors opened or closed till the shadows cast by the rod appear equally luminous .
The motor is stopped and the aperture read off .
Three or four readings for each current passing through Ln are taken , and then the current is altered , and a new set of three or four readings made .
The current is altered so that the light from Ln varies from extreme brightness to a dull red .
By the other plan the sectors shown in fig. 2 are detached from the motor , and card disks placed on A. A resistance , R , fig. 3 , has now Fig. 3 .
to vary to bring the light equal to the standard light diminished by the disks rotating between the lamp Lq and the screen with known apertures .
For the resistance at first we used a non-conducting tube , in which about forty spherical pellets of hard carbon were inserted .
At one end of the latter was a brass plate to which one terminal of the battery was attached , and at the other a screw was inserted , which was attached to the other terminal of the battery through the lamp .
This screw pressed the pellets together to any required degree , diminishing the resistance or increasing it as occasion required .
This answered fairly well , but not so well as would be desired , as the response to the screw was some1887 .
] On Photometry of the Glow Lamp .
what sluggish .
Mr. Varley supplied ns with one of his carbonised cloth resistances , which consists essentially of a series of square pieces of carbonised cloth more or less in contact .
The figure represents the one we had made for us .
The carbonised cloth is represented by C , fig. 3 , which fills the whole length from A to D when loosely packed .
At B is a plate to which T3 is attached , and which can be separated more or less from a fixed metal plate to which Tx is connected by the arm E , which is moved by the screw Sp At A is an insulated block carrying another plate to which T2 is attached , and A can be carried backwards or forwards by means of the screw S2 .
Eor some purposes the main current can be brought in at T3 , and leads be taken from T2 and T1 } thus forming part of a Wheatstone bridge .
When ouly one resistance has to be inserted , Tx and T3 , Tj and T2 , or and T3 may be used for connecting on to the leads with one pole of the battery and the ampere-meter .
It was in this way the resistance was used in the case in point .
The lamp Ln by the use of this could be raised from black heat to bright white , and a small turn of the screw altered the resistance very considerably .
We used two sets of sectors ; one pair enabled us to use an aperture from 135 ' to 90 ' , and the other pair from 90 ' to 0 ' .
The light from Iq was diminished by the first pair of sectors being placed on A ( fig. 2 ) , the current cc potential being noted .
The resistance of the curi'ent passing through Ln was then altered till the illumination of the two shadows on the screen appeared equal , the screw S ( fig. 3 ) being turned backwards and forwards , first one shadow and then the other being made to appear too dark .
By diminishing the oscillations the neutral point can be very readily arrived at , even though the colours of the lights may be very different ( see Bakerian Lecture , 1886 , " Colour Photometry , " by the authors ) .
The readings of Yq and Ajj were then read and noted .
The apertures of the sectors were altered , and the same operations gone through .
Ei'om observations thus made the curves were constructed , enabling the theory propounded to be tested .
I he grease-spot plan of photometry was arranged in a somewhat similar manner , Ln being on the opposite side of the screen SS , and the rod being abolished .
In this method the room has to be dark so as to admit of no reflection .
At first we were not prepared for any great exactitude with it , but finally we came to the conclusion that it was very reliable , a conclusion that Mr. W. H. Preece* came to when he constructed his photometric arrangements .
Having described the arrangements for taking the measurements , it remains to give the conclusions at which we arrived after making a large number of experiments .
* ' Roy .
Soc. Proc. , ' vol. 36 , 18S4 , p. 270 .
VOL. XLIII .
n Captain Abney and Major-General Festing .
[ Dec. 8 , Let W be the watts , c " current , p " potential , y , , intensity of light ; all other letters being constants .
In order for the curve of intensity to be parabolic ( i. ) W\#151 ; m = n*/ y , m being the number of watts at which the vertex of the parabola lies .
From the equations given in the paper already referred to ( ' Phil. Mag. , ' September , 1883 ) , where c = ap -(- 6p3|2 } the above equation ( i ) may be written\#151 ; p^(a + bpi)\#151 ; m = when p is fairly large this becomes\#151 ; ( ii ) p2\#151 ; h \#151 ; h*/ y approximately .
Similarly it may be shown that\#151 ; ( iii ) c2\#151 ; s = t*/ y approximately .
The following tables will show the application of ( i ) .
It must be understood that the measures of current and potential are not given in amperes and volts , and that as a consequence the watts are only represented by watts x a constant .
The first three tables show the exactitude of the method of measurement where the resistance is altered , and the fourth table the exactitude when the rotating sectors are altered .
887 .
] On Photometry of the Glow Lamp .
Table I.__A Wood house and Rawson lamp , changing the resistance and reading current and potential .
c. P. Watts x a constant .
Aperture of disk .
Calculated aperture ( shadow ) .
Calculated aperture ( grease-spot).# 13-4 38-0 509-2 180 180-23 181-6 12 -75 36-1 460 3 135 135 -20 134-2 12-5 35 5 443-8 120 121 -44 120 0 12-3 350 430-5 110 110-95 109 -4 12-2 34-7 423-3 105 105-47 105-5 11-9 33 8 402-2 90 90-11 91*1 11*7 33 1 387 3 80 79-99 80-1 11-5 32-5 373-7 70 70-34 71 -2 11-2 31-7 353-9 60 60-20 61-3 10 -9 30 -75 335 -2 50 49-40 48-7 10 6 29-9 316 -9 40 40-42 39-8 10-2 28-8 293-9 30 30 11 29-7 9-65 27-55 265-4 20 20-07 19-5 9-35 28-7 249-6 15 15-07 15-0 9-0 25 6 230-4 10 10-09 8-4 24 -2 203-3 5 4-75 1 m = 144 .
n = 27'2 .
Table II.\#151 ; Swan lamp-light measured by changing the resistance and reading current and potential .
C. P. Watts x a constant .
Aperture of disk .
Calculated aperture .
23-2 24-2 562-4 180 180-63 22-0 22-2 ' 488-4 133f 132 -71 21 6 21-6 466-6 120 119 -89 21-4 21-4 458-0 115 114-91 20*6 19-9 410 -0 90 89-14 20-2 19-2 387-8 79f 78-68 19-7 18-7 368-4 70 69 -88 19-2 18-0 345-8 60 60-37 18-6 17-2 320-0 50 50-00 17 -9 16-3 291 -8 40 40-25 17 1 15-3 261-6 30 30 -8 15 -9 13-8 220 -4 20 19-98 14 -3 12 0 171-6 10 10-11 13 -0 10-6 137-8 5 5-29 m = 50*52 n = 38*0 .
The following shows the readings in full of one set of observations :\#151 ; * This column shows results given by the grease-spot method .
The watts are not given , but merely the results , to enable a comparison to be made between the accuracy of the two methods .
t In these two the sectors were supposed to be set at 135 and 80 respectively , but after the set had been taken and the sectors stopped it was found they read as in the table .
u 2 Table III .
Current .
Readings .
19'3 18-5 17-0 16-8 16 5 15-6 15 2 14-5 13-7 12-2 10-6 19-3 17-1 16 -7 15-6 Mean .
Potential .
Aperture .
Readings .
Mean .
Watts x a constant .
Calcu- lated .
Set on the sector .
i 2 3 4 11-4 11-5 11 4 11-4 11-4 220 180 180 10 8 10-85 10-75 .
, 10-8 200 135-5 135 10-3 10 3 10 3 10-3 10-3 176 89-5 90 10-1 10-1 10-1 10-1 10-1 170 79-6 80 9-9 10 9-9 , # 9-9 163 71-7 70 9-75 9-75 . . .
, 9 -75 157 59-9 60 9-6 9-5 9-5 9-6 9-55 149 49-2 50 9 3 9-3 9-5 9-2 9-4 142 40-7 40 9-1 9-1 9-0 , , 91 132 30-0 30 8-7 8-7 8-65 8-7 8 -7 119 18-7 20 7-9 7-9 7-9 7-9 7-9 96-5 5-6 10 7-3 7-1 7-2 . .
7-2 76-3 0-14 5 71-5 .
11-04 .
a cp \#169 ; o p 254 Captain Abney and Major-General Fes ' 1887 .
] On Photometry of the Glow Lamp , The following is an example of measuring by using known currents and cutting off more or less of the comparison light by the sectors .
The observations have been given in full to show the deviation of individual observations from the mean :\#151 ; Table IV .
c. P. Watts x constant .
Aperture of disks to balance light .
Mean observed aperture .
Calculated aperture .
7-9 28 -1 222 90* 90 91 -2 7-75 27 -4 212 81 , 81 , 80 80-7 80-6 7-6 26-8 204 72-5 , 72-5 , 72-5 72-5 72-5 7 25 25-5 185 56-5 , 56 , 56-0,55,55-5 55-8 55-6 7-00 24-8 174 46 -5 , 47 , 46 -5 , 45 *5 46 " 4 46-6 6-7 23-8 159 37 , 35 , 36 , 36 36 -0 36-0 6-25 22-2 139 23 , 24 , 23-5 , 24 23-8 23-6 5-85 20-6 120-5 14-15 , 15 , 14-5 , 14 14-5 14-6 5*7 20-1 114 11-75 , 11-75 , 11-75 11-75 11-75 5-4 19-1 102 8 , 8 , 7-5 , 8 , 7-5 7-6 7-7 53 18-7 99 6-5 , 6-5 , 6-5 , 7 6-6 6-8 m = 53 .
n = 17'7 .
The foregoing examples will give an idea of the accuracy with which measurements may be made by either method , and of the exactness with which the parabolic curve is followed .
It seems that the photometry of incandescence lamps may be well carried out by measuring the watts .
It may be objected that each observation requires readings of the galvanometers , but this is avoided by the use of the formula given in the beginning of this paper .
Two observations of current and potential enable the constants to be calculated , and after that one galvanometer alone need be used ; by preference that one giving comparative volts .
The current is calculated from such a reading and subsequently the watts .
Mr. W. H. Preece , in his paperf already alluded to , came to the conclusion that the intensity of the light emitted from a glow lamp varied as the sixth power of the current .
This formula is fairly exact within limits , but it is obviously empyric , since where the current is small enough only to cause dark radiation it must fail .
Ihe example that he gives would require some slight rectification before it can be used as in the method given above ; since the small distances at which the candle he employed was placed from the screen make it necessary to apply corrections for the thickness and length of flame .
* The light was fixed , so as to balance as nearly as possible when the sectors were at their full aperture .
f ' Roy .
Soc. Proc. , ' vol. 36 , p. 270 .
256 Captain Abney and Major-General Festing .
[ Dec. 8 , Mr. Preeee 's table is as follows ; the last column is derived from the parabolic formula using C2\#151 ; sinstead of W\#151 ; m = aVy .
Table V. Distance of source of light from illuminated surface .
Equivalent degree of illumination .
C urrent in lamp .
C6 x 15-994 .
Parabolic formula .
0 -50 feet 64-000 1-260 64 -000 64-00 0-75 " 28-445 1-100 28 -335 32-83 i-oo " 16-000 0-959 12-442 16-00 2 -00 " 4-000 0-790 3-888 4-41 3 -00 " 1-778 0-690 1-726 1-78 4 -00 " 1-000 0-651 1-217 1-00 It having been shown that the parabolic formula applies to visual measures of an incandescence light , it appeared that the same ought to hold good for the total light which is photographically active .
These rays may be taken to lie between the blue and the extreme ultra- Fig. 4 .
1887 .
] On Photometry of the Glow Lamp .
257 violet of the spectrum , and consequently the vertex of the parabola should lie further towards the blue of the spectrum than it does with the visual rays .
The method of testing was as follows .
An ordinary dark slide , A , carrying a sensitive plate , was placed in grooves , CC , attached to a board , D , against which the slide ( when its front B was drawn out ) could be raised or lowered as occasion required by means of a rack and pinion motion working by the handle D. Fro .
5 .
E is a projection in which a slot , H , is cut , and through which a card having a square aperture , K , can slide .
K can be covered by means of a cardboard screen .
The bottom of the plate in A is first brought opposite the aperture K , which is placed opposite the number marked 1 on E. The lamp is placed 4 feet away , the volts and amperes noted , and exposure is given to the small square of the plate seen through the square in K for any time which may be fixed upon .
The slide A is lowered , so that a fresh portion of the plate is brought opposite to K ( K being covered up ) a different current passed through the lamp , and another exposure given , and so on .
When the top of the plate has been arrived at by the motion of A , the card F is moved till K is opposite 2 , and the same procedure repeated .
Six to ten exposures can be made in the same row .
When the second row is exhausted , K is placed opposite 3 , and 258 Captain Abney and Major-General Festing .
[ Dec. 8 , such a current is passed through the lamp that it emits a medium light .
A time scale is then made by giving different lengths of exposure at each movement of the plate in its last half .
The plate is then taken out of the slide , and the images developed .
By this means both a time scale and a measure of intensity for different temperatures are on the same plate in the shape of squares of different density of deposit .
When the negative is dry it is placed in an apparatus which works on the principle of the optical lantern , and is described in the paper written by one of us , " Atmospheric Absorption of Sunlight " ( ' Phil. Trans. , ' 1887 ) , and the density of each square measured .
The " intensity " measures are then compared with the time scale , and the value of the intensity calculated .
From these values can be determined if the curve of intensity and watts increases parabolically .
It might be objected that increase in intensity is not convertible into " time of exposure .
" Careful experiments have been made as to this , and for the range of time which is comprised in the seconds of exposure given no appreciable error ensues .
The following is an example of an experiment conducted in the above manner:\#151 ; Table VI .
Time Scale .
No. of aperture .
Exposure given to portion of plate .
Light transmitted through developed plate .
secs .
i 5 550 2 10 47*5 3 15 39-5 4 20 330 5 25 27-5 6 30 23-2 7 35 20-0 8 40 18-1 9 45 16-0 10 50 141 11 55 13*2 12 60 12 -4 Bare glass = 56 .
PL .
CD \amp ; % *4 ^ O CD it ?
d 2 .
p cr cd sr g \lt ; 1 \amp ; \#163 ; gf CD O O. ^ M .
P cf-CfQ P-1 ?
J \#174 ; P g g. . .
cr th ^ 05 o \lt ; ?\gt ; ct .
\amp ; S I ?
\#171 ; 0Q p CD GQ Table YII .
No. of Lamp .
Exposure in seconds of small squares .
Light trans- mitted .
Equivalent time exposure .
Mean .
Reduction to one Calculated By least exposure .
Current .
Potential .
Watts x constant .
minute 's exposure .
intensity .
squares .
1 15 0 37 -5 562 15 21-5 66\ 264 264 264-8 2 J3 215 66 J DO 3 14 T 35-3 498 30 16-25 881 88-5 1 \gt ; 7*7 185 176-2 4 16-00 89/ \#177 ; 4 4 5 13-2 33-3 440 30 25 561 iiq 114-5 1111 6 24-5 57/ do o l\#177 ; o 7 12-6 32 1 404 30 32-5 401 A 1 oo 79 -2 78-3 8 31-5 42 J 41 o2 9 11 '6 29-8 346 60 35 36 1 O JC or* .
A 37-4 10 37 34/ 35 Ot\gt ; oo U 11 11 1 28-6 317 120 31 -5 421 ol Ol 99 .
1 12 31-5 42/ 42 21 21 22 1 13 14 10 3 33 26 -9 277 180 33 34-5 401 37/ 38 *5 12-8 6*8 8-4 15 33 9-4 33 24-8 231 240 42-5 25 1 6-2 16 42-5 25/ 25 0*9 33 33 33 33 17 9 2 24-4 224 360 39 311 18 39 31 J 31 5 " 2 0 31 33 33 33 19 8-6 22-9 197 480 48 171 2-25 20 m 19 } 18 0 33 33 33 1 33 w = 235 .
n = 20 -08 .
On Photometry of the Glow Lamp .
259 Captain Abney and Major-General Festing .
[ Dec. 8 , dimensions ) was constructed .
Take Nos. 1 and 2 , Table VII , the exposures given to each were the same , viz. , 15 seconds .
The light transmitted through each square was read off , and found to be identical , viz. , 2T5 .
In the scale diagram the abscissa of the curve having the ordinate 2T5 is 66 , which is the equivalent time exposure , which is also a measure of the intensity .
If the exposure had been prolonged to 1 minute it would be equivalent to an intensity of 264 ( since the measured exposure was only 15 seconds ) .
The other intensities were calculated in a similar manner .
The intensities 21 and 264 were taken as points on the parabola , and the intermediate intensities calculated using the formula W\#151 ; m = nVy .
Fio .
6 .
Time Scale Diagram .
\#151 ; It will be noticed that in the above diagram the abscissae are marked as square or relative exposure .
A reference to the time scale table will show that the exposure = number of square x 5 .
For simplicity 's sake the number of the square has been taken as the unit of time .
The following is another example of the photographic measurement .
1887 .
] On Photometry of the Glow Lamp .
Table VIII .
Time Scale .
No. of exposure .
Time .
Readings of density .
15 180 7-5 16 160 8-5 17 140 9-5 18 120 12-0 19 100 17 0 20 90 20-7 21 80 25*1 22 70 30 -0 23 60 375 24 50 46-5 Table IX .
1 No. of Lamp .
Exposure Light transmitted Equiva- lent exposure .
Mean .
Reduced to one Calculated intensity .
exposure .
Current .
Potential .
Watts x constant .
in seconds .
through negative .
minute 's exposure .
1. .
8-6 32-7 281 1K 7 Ol .
\#163 ; 1 Q/ f 1 ( for 15 sec. ) 186 2 h lOi !
1 OQ f 744 736 55 99 99 4 LiL D oo J 3 8-0 30-2 242 q n ( for 15 sec. ) 134 4 Ou 1 K not re\amp ; ci i n a Die 134/ 536 529 55 99 99 JLD 5 7*7 29-0 223 Af\ not read A able " 1 144 J 6 on ( for 20 sec. ) 432 432 ) ) 99 99 y 144 7 6-6 24-9 ftn 8 12 *5 1681 118/ 8 AOv Otr / tn ( for 60 sec. ) 176 99 99 j ) *\#177 ; U 172 172 9 6*1 22-9 1 QQ fin 10-5 33 -5 130*1 10 joy oU on ( for 60 sec. ) 130 130 99 99 99 oU 65/ 130 11 5 *4 20-0 99 i no 120 60 9-75 31 138*1 70/ 12 99 XU2 55 ( for 120 sec. ) 139 69-5 69-2 13 4 *4 16-2 *71 600 300 9-5 29-5 1401 ( for 600 sec. ) 14 / 1 14 1 15 -2 99 99 55 71J 141 262 On Photometry of the Glow Lamp .
[ Dec. 8 , 1887.1 The Detonating Bolide November 20 , 1887 .
The above show that the parabolic form seems to be followed , but owing to the want of absolute uniformity in all parts of a photographic plate , and that errors may arise from want of exact exposure , and , again , from reading the densities , the values obtained are not so accordant as those taken by the visual method .
V. " On the Detonating Bolide of November 20th , 1887 .
" By G. J. Symons , F.R.S. Received December 8 , 1887 .
Shortly after November 20th it was generally reported that an earthquake shock had been felt in the South Midland counties of England , and the author began to collect and examine the facts .
It appeared that the records from Oxfordshire , and the western stations generally , indicated that much louder sounds were heard there than at the eastern stations , e.g. , Essex and Cambridge .
The author thought that , although the phenomenon had been almost universally ascribed to an earthquake , it was more probably due to an explosive bolide , and on receiving from one of the local scientific societies a request for assistance in tracing the shock , the author suggested the alternative explanation .
Mr. Fordham has subsequently written to say that he has already found one person who saw the meteor from Hertford , which he describes as " a brilliantly luminous body travelling across the sky from N.E. to W. " It is further stated that a portion of the meteor was seen to fall from the main body .
Considering that the morning , as shown by the records of the Royal Meteorological Society , was both misty and cloudy , and that at the hour at which it appeared , Sunday morning , 8.20 a.m. , there would be broad daylight , it is improbable that many persons saw it .
Judging by the descriptions of the noise , as well as by the path roughly indicated by the Hertford observation , it seems likely that it ex-

