104 PLANT PHYSIOLOGY. 



over the tip of the shoot. Fasten, by a small drop of glue, to one of 

 the smaller wheels the end of a silk thread thoroughly waxed (to 

 lessen changes in length through hygroscopic changes) and make 

 several turns around it. To the groove in the larger wheel fasten 

 also a similar thread which makes a turn in the opposite direction. 

 Attach the free end of the first thread to the tip of the part of the 

 plant to be measured, and the end of the other to the recording-pen, 

 which can be made to press against the cylinder by giving it several 

 turns in the wrong direction. Thus arranged, the wheel will turn as 

 the plant grows, and the pen will trace a spiral line on the revolving 

 cylinder. The exact amount of growth through the 24 hours can 

 then be read off for every hour by placing a millimeter scale along 

 the cylinder from the starting-point downward (or on the paper re- 

 moved from it), on which of course the record appears magnified in 

 proportion to the comparative diameters of the wheels used. It is 

 well to try one record through the 24 hours in a thermostat kept at 

 constant temperature, and another in the ordinary fluctuations of the 

 greenhouse, temperature being read from the thermograph. Partic- 

 ularly good for the study of growth are vertically and rapidly 

 elongating flower-stalks, such as those of the hyacinth or grape- 

 hyacinth. The paper is to be removed from the cylinder for 

 preservation, and a curve is to be plotted from it. 



The advantage of the quadruple over a double wheel (such as is used 

 in Fig. 24) is that it allows of different degrees of magnification, and 

 hence makes the apparatus available for other purposes where continuous 

 records are needed. Owing to the leverage exerted by the pen when 

 placed on the largest wheel, it tends to exert a considerable pull upon the 

 plant, thus affecting its growth ; hence the pen should be made as small 

 and light as possible. For exact work, it would be better to use a 

 counterweight to balance the pen (using a close coil of copper wire 

 which may easily be made the same weight as the pen), and to make the 

 wheel turn by a small weight placed on the thread coming from the plant 

 and passing over (and once around) the smaller wlv.-el. 



(The cylinder should be of hardwood, since weight is not a drawback 

 in this case, and tJie holes for setting it on tlie clock-spindle will not 

 wear loose as with soft wood. When set revolving on its a.vle, tJie 

 wJieel should come to rest indifferently at any point. If it does not, 

 small pieces of paper can be gummed to the lighter side tint il it bal- 

 ances. T/ie clock must be wound and a new record commenced every 

 twenty-four hours, but a vigorous plant will cover the cylinder with 

 its magnified record in that time. For sonie purposes an eigJit-day 

 clock would be better. Watering the plant introduces an error, as it 

 swells the soil and lifts the plant a trifle' ; hence the watering should 

 be done just before the new record is begun. Some practice will be ncc- 



