550 LECTURE XXXII. 



been the subject of investigation. Christopher Jacob Trew, even so long ago as 1727, 

 published long-continued daily measurements on the flower-scape oi Agave americana, 

 combined with observations on the temperature and weather ; but it was not till a 

 hundred years afterwards that Ernest Meyer (1827) and Mulder (1829) gave a new 

 impulse to investigation in this direction, and they were then followed by Van der 

 Hopp, De Vriese (1847-48), and others. The important questions, however, were 

 gone into more thoroughly by Harting (1842), Caspary (1856), and Rauwenhofif 

 (1867). 



'These observations, industriously and perseveringly carried on, led to no definite 

 answer, nor even to the establishment of an actually useful method, and a careful survey 

 of them shows that scarcely any two observers came to the same conclusion, and that 

 the discovery of any causal relation of growth in length to temperature and light was 

 in fact impossible, since on the one hand the questions to be answered were not put 

 with sufficient clearness and definiteness, and on the other, the probable sources of 

 error, and, accordingly, the difficulties of accurate observation, were left more or less 

 unnoticed. Meanwhile there appeared another series of publications simply of 

 repeated measurements of lengths, no regard at all, or not sufficient, being paid to 

 the external conditions, whence although some idea was obtained of the continual want 

 of uniformity of growth during different days and at different times of the day, the 

 causes of these were not registered: some observers, in fact, confined themselves 

 to merely demonstrating the difference between diurnal and nocturnal growth, 

 not reflecting that ' day ' and ' night ' signify to the plant different and very variable 

 complications of conditions of growth, and that such a mode of stating the question 

 cannot possibly lead to the discovery of causal connections, so long as the individual 

 factors which are contained in the ideas day and night (so far as the plant is con- 

 cerned) are not known. In this sense the contributions of Seitz, Meyen, Martins, and 

 Duchatre, for example, are useless for our purpose. 



'I have, though with considerable interruptions, been engaged since 1869 with 

 the completion of more exact methods of observation, and have attempted to classify 

 and render clear the questions to be answered.' 



This was done by means of the following considerations. ' With few exceptions 

 the majority of observers of growth in length have selected parts of plants which are 

 remarkable for a very considerable growth in a short time. The huge flowering 

 stems of the Agaves especially afforded repeatedly, on account of their rapid growth, 

 the inducement to the making of such observations: such objects were selected 

 because observers were satisfied simply to measure the growth in length with a 

 rule laid directly on the part of the plant, observed. But even granted that in the 

 case of quickly growing plants sufficiently exact measurements are to be made in this 

 manner in intervals of time of from one to several hours, still, other evils make 

 their appearance of which I will particularise two only. In the first place, plants 

 which grow so rapidly that even only four to six sufficiently exact measurements 

 daily can be made, are rarely to be obtained : one is at the mercy of chance, 

 and a methodical and connected series of observations can scarcely ever be com- 

 pleted. In the second place such plants (e. g. Agaves, Musace^, Victoria regia) are 

 usually so large that it is necessary to undertake the observations in greenhouses or in 

 the open air, and therefore under conditions where they are exposed to very large and 



