54 
(16) The absence of any barographic sign of a great and 
instantaneous decrease of pressure. 
(17) The presence of a temporary rise, accompanied by a 
sudden jerk, the final result, at least, being upwards, 
of 0 02 at York, of O’IT at Cincinnati. 
(18) In all five cases the barometer at the time was near, 
or at, the end of a rapid fall. This is, of course, 
in close connexion with such storms occurring in 
the S.E. octant, if the depression is following the 
usual S.W. to N.E. course. 
We may note that, not only is there no absolute proof as yet 
that tornadoes blow along the ground in whirls, but, also, none 
of the stories of corks leaving empty bottles, &c , from any 
sucking action, causing a partial vacuum, have as yet been 
confirmed. Again, houses are said to have burst outwards 
because of a sudden diminution of the external pressure Such 
cases, however, are explained by the wind getting in on the 
exposed face and hence, naturally, bursting the others outwards. 
There is, plainly, some relation between the tliunderstorm and 
the wind-rush, but our present knowledge does not appear to 
be sufficient to say what. The latter, it will be noticed, moved 
across the face of the former. In this way it differed from the 
sudden gusts, wdiich sometimes accompany such storms. The 
thunderstorm itself is of the type to which squalls belong, as 
was pointed out earlier. These, again, differ from the great 
cyclonic storms, which oiten affect half or all our country at a 
time, when the wind blows inwards, spirally, towards the centre 
of depression. 
The thunder-cloud, it must be noted, moved almost at right- 
angles to the wind-circulation of the main cyclonic-storm, just 
as the wind-rush was perpendicular to it. Hence the wind-rush 
was parallel to the direction of the general wind of tlie 
afternoon It may possibly have been a portion of it intensified, 
in some way, by the passage either of the thunder-cloud or of 
the unknown cause by which thunder-clouds are produced. 
J. Edmund Clark. 
