64 Dr. van Rijckevorsel on Irregularities in the Yearly 



although not much, higher than the August one. This is an 

 indication of the southern origin of this July maximum. May 

 not this be an explanation ? Lnte in June Europe is invaded 

 from a point off the Scottish coast by a strong minimum, but 

 a week or so later by a strong maximum from the south or 

 southwest, let us say tropical Africa. Therefore it is probable 

 that the fight, so to speak, of these two features, in the first 

 place causes the distribution just spoken of, and that the 

 curious particulars w r hich accompany it must find their expla- 

 nation in local circumstances. 



A wild hypothesis ? I know it is. But 1 give it in order 

 to show how vividly some meteorological problems are put 

 before us by this method, and how a path is shown at the 

 same time to their evolution. For can there be much doubt 

 but that, as soon as we shall be possessed of more data, not 

 about temperature only, more and more of these problems must 

 make great progress towards a solution ? One curve for 

 magnetic declination or for rainfall may throw a ray of light 

 on an intricate point which at once solves it. 



Another argument to the same purpose. There will be 

 seen, not without some effort perhaps, a very slight minimum 

 between the 10th and 15th of May in the three upper curves 

 (PI. II.) . There are other features of the same importance, of a 

 greater one evem There need be no fear of my discussing them 

 all, for the very plausible reason that I do not know anything 

 about them. They may either prove interesting some day, 

 or vanish completely, when meteorologists shall be able to 

 discuss longer series of observations. But this one is worth a 

 moment's notice, for it is really the mark of the so-much 

 discussed u Ice-saints." You will notice a trace of it in the 

 three upper curves, not in the two lower ones. (The rainfall 

 is too doubtful to be quoted here.) Well, these two lower 

 curves belong to Greenwich, and you will see that the hand- 

 mark, if I may express myself thus, of these cold saints 

 tends to vanish in the British Isles. The temperature-curve 

 for Greenwich indeed shows it still, that for Valencia perhaps, 

 but certainly in Scotland it is not visible. Here again the 

 southern, or southeastern, origin of this phenomenon, which I 

 hinted at at Toronto, is confirmed by other curves than those 

 for the temperature. 



As hinted at before, some of these curves leave for the 

 moment a doubt. It will be seen that between July and 

 November I left in some of the vertical lines a break in the lower 

 regions of my diagram. In nearly every case it is only the 

 barometer that is at fault. I already have stated the cause: 

 the series of 19 years which I used is not long enough. I 

 regret that 1 have been unable to be ready in time with a 



