A464 TRANSACTIONS OF SECTION A. 
large to show that some closer approximation is desirable, and a reconsideration 
of the process of dealing with cyclonic depressions as centrical distributions is 
a first step in that direction. 
The ‘average’ cyclonic depression is certainly not the most frequent—possibly 
it has never existed—and the typical cyclonic depression is of very rare occur- 
rence. Further, it is only for the distribution of pressure, and to a certain 
extent for winds, that any centrical symmetry can be claimed. ‘Temperature, 
clouds, and rainfall have no symmetrical distribution with regard to the baro- 
metric minima, 
The classical method, indeed, has always recognised the existence of deviations 
from symmetry, and has regarded them as ‘secondary depressions’ with certain 
recognised consequences. 
The report by M. B. Brunhes, Director of the Observatory of the Puy de 
Déme, upon the results of the competition in weather forecasting in connection 
with the exhibition at Liége in 1905 has brought into prominence two methods 
which lay great stress upon the importance of the deviations from symmetry, and 
constitute, as M. Brunhes says, a definite step towards a second approximation in 
forecasts of weather by means of synoptic charts. 
The two methods are by M. Durand Gréville, of Paris, and by M. Guilbert, of 
the Meteorological Society of Calvados, respectively. The paper gives an account 
of these two methods and illustrations of them by means of charts specially 
constructed for the purpose. 
M. Durand Gréville’s contribution consists chiefly in recognising the existence 
of disturbances of the smooth course of the isobars surrounding a cyclonic centre 
extending in a long narrow band from the centre or from near thereto. The band 
consists generally of an exaggeration of the fall of pressure along the front of the 
band (couloir de grain) and a more rapid rise following it (ruban de grain), and 
then a resumption of the rezular march of the isobars until a second ligne de grain 
arrives. While the rwban de grain is sweeping over a district in a manner similar 
to the march of a wave-front, changes in the direction and force of the wind, and 
rainfall, occur with varying degrees of intensity along the line. In some localities, 
where conditions are favourable, line-squalls or thunderstorms are developed. In 
order to exhibit the existence of these lignes de grain M. Durand Grévyille adyo- 
cates the drawing of isobars for intervals of 1 millimetre instead of for 5 milli- 
metres, as is usual in Continental charts. 
The characteristics of a ligne de grain are illustrated by the diagrams pre- 
pared by Mr. Lempfort for his paper on the line-squall of February 8, 1906, read 
before the Royal Meteorological Society in May 1906; and the special maps pre- 
pared for further illustration are tbose for June 22-23, 1900, November 1-2, 1893, 
and December 7-8, 1899. In these the pressure is expressed in degrees of pressure 
(2000 dynes per sq. cm.) and isobars drawn for intervals of one degree of 
pressure (15 mm.), as explained to the Association at Cambridge in 1904. 
Maps were constructed on a similar plan to illustrate M. Guilbert’s contribution 
for August 31-September 1, 1905, September 4-5, 1905, and February 19-20 -21, 
1892. 
M. Guilbert’s method depends upon the comparison of the actual winds, as 
recorded on the map, with ideal or normal-winds as computed from the distance 
apart of the cousecutive isobars. 
The method of calculation of winds employed in the Meteorological Office in 
accordance with the formula 
y= A(2@V sind + V?cotp/R) 
is considered, and M. Guilbert’s estimate of normal winds compared therewith. 
Examples are given in the maps already specified of M. Guilbert’s application of 
the vector-deviation from the normal winds to determine the localities of the rise 
and fall of the barometer in the ensuing twenty-four hours. 
M. Guilbert’s method is summarised by the statement that where the vector- 
deviation forms an anticyclonic system barometric fall is to be expected, and 
where the vector-deviation is cyclonic a rise of the barometer is to be expected. 
