1909.] Vegetable Assimilation arid Respiration. 55 



ably less than 0'5 per cent, of areas as small as 20 sq. cm. : the other 

 methods are at least as accurate as this. 



All the errors of the templet method can be reduced to a minimum by 

 modifying the form of the templet and of the cutting instrument. 



By means of the rotating ■punch small discs may be cut from between the 

 veins. Since, however, this method is not a very rapid one it should not be used 

 for leaves like those of Heliantlms anmius, which are liable to rapid shrinkage. 



The planimeter method can be used for entire half-leaves, or for pieces cut 

 from between the veins ; great care is necessary, however, especially in the 

 latter case, to avoid shrinkage while taking the photographic print. 



The stamping method completely eliminates all shrinkage errors ; by it the 

 dry weight method is made as simple as possible, but its use is restricted to 

 fairly smooth leaves. With both the templet and stamping methods areas as 

 small as 10 sq. cm. can be safely used ; but the disc and planimeter methods 

 are more suitable for very small leaves. 



4b. Since dry leaf material is often extremely hygroscopic, it should be 

 dried in a current of dry air and weighed with great care to exclude moisture. 



5. To sum up, the main errors involved in the dry-weight method are due 

 to (1) the shrinking of leaves in area during experiment, and (2) their lack of 

 symmetry in respect of dry weight per unit area. Of these errors, those 

 from shrinkage can be eliminated ; those from asymmetry, on the other 

 hand, must always be reckoned with. Asi/iametrij tests must therefore always 

 form, an integral part of dry weight experiments ; especially as any appreciable 

 reduction in experimental errors will also appear in these tests. 



The relative magnitiide of the total error may be rediTced by increasing the 

 time of experiment and so distril)uting it over a larger number of hours. 



The reed increase of dry weight in a period of 5 hours or more can he 

 determined with most leaves correct to the nearest 'milligramme per sqtiare 

 decimetre per hour hy using appropriate methods* With shorter experi- 

 ments this degree of accuracy cannot be reached except with very thin leaves. 



Thus, the dry weight method is capable of yielding useful results, for rates 

 of increase greater than 2 milligrammes per sqiiare decimetre per hour. It is 

 not to be compared for accuracy with gasometric methods, and can never be 

 expected to afford a means of studying the more refined questions of pure 

 physiology ; but the accuracy_, obtainable will allow of the approximate 

 solution of a number of interesting problems, and will be of service in 

 ecological comparative studies of assimilatory activity. 



* This applies strictly only to broad leaves. For narrow linear leaves special nietliodS 

 would have to be devised, and for many very small leaves the half-leaf method is 

 obviously unsuitable. 



