4-8 Dr. Ballot on the importance of deviations from the mean 



section of the cylinder; the distance between the two circles 

 will be proportional to the height of the barometer at that 

 place. This will be true for all places; and the height of the 

 letters or signs (a) (b) noted at the respective sections in the 

 surface of the cylinder would indicate the height of the baro- 

 meter at those places. If we imagine our eye in the axis of 

 the cylinder, and we project those supposed points (a) (b) 

 on the horizontal plane, or what is the same thing, if we apply 

 the method of descriptive geometry to this, the greater or 

 lower height of the barometers at (a) (b) will be represented by 

 points, «, /3, in circles concentric with the first, but more or 

 less removed, whose distance from the first (the difference of 

 the radius) will be proportional to that height. So the first 

 circle may represent a very low barometrical height, as730 mil- 

 limetres, and each millimetre that the barometer in (a) or (b) 

 is higher may be represented by one or two millimetres' di- 

 stance of the circle (a) or (/3) from the first. But we can go a 

 step further; we can divide those circles from the common 

 centre point into a number of sectors, thirty sectors 1 suppose : 

 in one sector the letters may signify the height of the baro- 

 meter at the 20th of October, in another at the 21st, and so 

 on ; every sector may be applied to one day. So we shall 

 have the barometrical range at every one place, and the differ- 

 ent states at the different places simultaneously, equally well 

 and clearly represented, while the map within the circle indi- 

 cates the distribution with respect to the surface of the earth. 

 If, by chance, the points (a) in the different sectors lie in a 

 circle, it indicates that the barometer at a has not varied du- 

 ring the month; if they lie in an ellipse, then there have been 

 two highest and two lowest states; we see immediately those 

 heights and the dates. So, as to the simultaneous comparison 

 of the barometer heights, the eye catches immediately the let- 

 ters that are higher or lower, and we see on the map in the 

 middle the line of places where the barometer was at the high- 

 est or the lowest. If there actually is a wave, then it is clearly 

 exhibited thus to the eye; if the eye has difficulty in discern- 

 ing the wave, then probably there is no wave, or perhaps that 

 wave must have been marked out by the foregoing or following 

 days. 



It is even possible to go a step further, and, for example, 

 to represent at the same time the thermometer state by the 

 distances at which we set the letters («), (b) (the signs of 

 the places) from the radii which divide the circle; the more 

 on the right, for example, the higher the state of the ther- 

 mometer. It is evident, however, that the more variations 

 we exhibit, the more is lost in simplicity. We could have 



