It may, perhaps, be a matter of surprise, that I should have undertaken these three separate tests, and left them in their initial stages, instead of working persistently at one in particular, and thus, maybe, putting the time to better use. The reason was the old and troublesome one which was always cropping up and causing me no little worry: _Lola's interest must not be allowed to flag_. In the course of a fortnight or three weeks, for instance, I have not dared to embark on more than _one_ test, not even continuing that one for as many as five consecutive days. This is why the three tests, above narrated, followed close one upon the other, while I took care to turn Lola's attention from them in between, making her go over all sorts of sums and spelling exercises. Should I have persisted in fixing her attention I should only have defeated my true object, and made her stale for future undertakings. In fact, I only engaged in these three, by way of giving a greater sense of _completeness_ to the idea, and also in order to fire the ambition of others embarking upon work of a similar nature. FORECASTING THE WEATHER On 2 May, 1916, at a season, therefore, when farmers are generally somewhat exercised as to the coming hay-harvest, and may well wish they had some contrivance--or knew of some method whereby they could ascertain, at all events, a few days in advance what the weather is going to be, a thought flashed into my mind. At first it raised a smile, it seemed so ridiculous and impracticable, yet there could be no harm in trying. I knew that most animals, such as birds, game, etc., sensed the approach of rain at least several hours before it began to fall. But the subject is one that has not yet come sufficiently under notice, so that we do not know whether they may not sense the atmospheric changes over an even longer period. We humans are not in a position to discover how animals come by their knowledge, we can only conclude that Nature has equipped them with more delicate "chords," so to speak, and that upon these highly strung chords she can sound a warning of her impending changes, since these, our humbler brethren, stand in more imminent need thereof. It is common knowledge that animals sense earthquakes long in advance of the actual shock, and this can only be accounted for in some such way. At the time of the earthquake in 1912, Rolf, at Mannheim, crept into a corner _several hours_ before it took place, and on being questioned, replied: "Lol hat angst, weiss nid vor was." (Lol is frightened; doesn't know at what.) It was quite useless trying to get further particulars as to his fears, for an earthquake was an entirely new experience to him; at a repetition of the event his remarks would, doubtless, be of greater interest and importance. Now as the weather is a matter that concerns animals, and with which they are also familiar, I determined to see how far I could get with Lola on this subject. So I taught her as follows: