254 THE JOURNAL OF BOTANY 



the flowers close, I find a note that, at about 5 p.m. on 6th June, 1882, 

 a dull rainy day, 1 was scarcely able to lind certain plants growing 

 among long grass in a meadow near the same place, though 1 had 

 seen them fully open and very obvious on the morning of the previous 

 day, which, though fine, had been dull and cool. 



In July 1882, when spending the summer at St. Moritz, in the 

 Ober-Engadine'(lat. 4G° 30' N., long. 9° 50' E.), I had a good oppor- 

 tunity of making systematic observations; for I found several plants, 

 growing among the grass on the mountain-side, within a few feet of 

 the door of my hotel, which had a southerly aspect. At the date 

 indicated, the flowers of this species would have been over, of course, 

 for some weeks in England, where the usual flowering time of the 

 plant is late in May and early in June ; but in the Engadine, at an 

 elevation of about G000 feet, they were still out fully. 



Broadly speaking, my observations (which have been confirmed 

 casually on later occasions elsewhere) went to show that, in 

 favourable weather conditions, the flowers always opened before 8 a.m. 

 and closed again before 12 noon. I confess I was never about early 

 enough in the morning to see the flowers in question before they 

 opened ; but I believe they did so as a rule shortly before 8. On the 

 19th July, a very exceptionally hot bright day, I noted specially that 

 the flowers were already open before 8, and that they had closed"a^ain 

 almost completely before 12 noon. On the following day, the 20th, 

 which was similiar in regard to weather, the flowers opened about the 

 same time, but had closed soon after 9. I was puzzled by this very 

 early closing, for which, at the time, I could see no reason. Perhaps, 

 however, the plant had some sort of intuitive forewarning (so to 

 speak) that the weather was about to change; for it did so completely 

 on the following day, the 21st, which was cold, dull, and windy, with 

 rain in the evening. On this day, the flowers which I found were 

 open at 8.30 had closed almost completely by 9.15, and did not open 

 again during the day. Yet, on the next day, the 22nd (which, 

 though warm, was overcast, with slight rain), the flowers reverted, 

 for some inscrutable reason, to their normal hours and remained open 

 till nearly 12. On the two following days, the 23rd and 24th, which 

 were both dull, cool, and rainy, the flowers again remained open until 

 nearly 12. There was nothing, so far as I could see, to explain this 

 change, unless, perhaps, it was to afford the flowers the greatest 

 possible chance of securing pollination before they went off altogether. 

 At any rate, at this time, most other flowers of the species were already 

 over and consecpiently my chance to make further observations came 

 to an end. 



The foregoing observations on T. pratense seem to show, so far as 

 they go : — 



(1) That, at Upsala, the open hours of its flowers occur some 

 three or three-and-a-half hours earlier than they do in the valley of 

 the Inn, some 13 degrees farther south and some six degrees farther 

 west, where the observations of both Kerner and myself were made. 



(2) That, while its flowers have their normal open-hours, their 

 hours of both opening and closing are liable to be varied considerably 

 (especially, it seems, those of closing) by the prevailing weather 



