20 6 PLANT PHYSIOLOGY 



THERMOSTATS. Of these, instruments for maintaining a constant tem- 

 perature regardless of external variations, many forms, from large rooms 

 down to small ovens, and of all degrees of precision, have been developed, 

 especially in recent years for bacteriological purposes. There is a firm in 

 Berlin, HERMANN ROHRBECK, making a specialty of their manufacture. 

 Forms constructed especially for certain botanical purposes have been 

 described by PFEFFER (Berichte der deutschen botanischen Gesellschaft, 

 13, 1895, 49) and by JOST (Botanische Zeitung, 55, 1897, 25). But all of 

 the above mentioned are dark, or at least lighted only from one side. The 

 first lighted form was that of SACHS (JahrLiicher fur wissenschaftliche Botanik, 

 a, 1860, 338), consisting essentially of a metal box, with a double water- 

 filled wall, fitting outside a flower-pot which is covered by a bell jar, 

 the whole heated by a lamp below; and this, modified in details, is figured 

 in his "Lectures," 278, and with the addition of a regulator in PFEFFER'S 

 "Physiology," 2, 83, though the proximity of gas-fumes in this form might 

 well be a great injury. This instrument is of course a uniform thermostat 

 giving only a single temperature, and is usable differentially only by employ- 

 ing several at different temperatures, or (less satisfactorily) by changing 

 its temperature from time to time. 



Differential Thermostats, viz., those giving a series of temperatures, and 

 these alterable at will, have been constructed for special purposes. There 

 is one made by LAUTENSCHI.AGER of Berlin, and another by PAUL ALTMANN 

 of Berlin. The latter has a series of ten chambers, cooled by ice at one end 

 and warmed by water-pipes and a gas-jet at the other. But the chambers 

 are dark and the instrument is very costly. I have, however, myself devised, 

 and am using, a much simpler and lighted instrument, which will later be 

 obtainable among my normal apparatus (Fig. 56). It consists of a copper 

 trough separated, by removable partitions, into ten compartments, covered ty 

 glass boxes; it is heated at one end from a copper box warmed by gas con- 

 trolled by a regulator, and is cooled at the other end from a copper box in 

 which cold water is kept circulating. A proper arrangement of pipes con- 

 ducts the gas-fumes completely away from the vicinity. The temperature 

 of the chambers grades evenly from end to end, and may be made to differ 

 any desired amount. A substitute could be improvised no doubt, especially 

 if fewer chambers are used, from a bar of metal resting at one end upon 

 steam-pipes, and cooled by running water or a box of ice at the other end, the 

 pots being covered by large tumblers or bottles. A rrake-shift method, 

 often employed in elementary work, consists simply in placing the pots in 

 different positions known to differ in temperature, but this is so liable to 

 introduce differences in other conditions also as to be very unreliable. 



Where it is impracticable to run the instrument overnight, it should 

 be placed in the coolest available safe place in order to check the growth 

 uniformly throughout. 



The graph of growth given by the earlier experiment shows 

 not only the fluctuations we are now studying, but also a 

 remarkable general curve, illustrating the grand period of growth, 



