INCREASE IN SIZE. 105 



essary to secure a good pen, but a few trials will give success. The 

 capillary tip of the pen sJiould be made smooth by rubbing it on a fine 

 whetstone or piece of ground glass, on which it should be rubbed at suc/i 

 an angle that the point will rest at right angles to the cylinder. Or, 

 with great caution, it can be smoothed in the gas-flame. These pens 

 can be made to write with any desired degree of fineness, even down to 

 a scarcely visible line, the finer the point is drawn out the finer the 

 line, but such fine pens need a smooth paper. It is better to make a 

 number of pens and cJioose the best. It may be filled by connecting 

 the large end with a pipette by means of small rubber tubing ; by forc- 

 ing a little air from the pipette with the tip of the pen in the ink 

 as the bulb is allowed to expand, the ink will enter the pen ; a few 

 drops are sufficient (or the bit of rubber tubing alone may be used on 

 the same principle}. If the brass loop (made from a brass paper- 

 fastener} grasping the pen is made a little small, the pen may be forced 

 into it and be tightly gripped. The thread may be attached to the 

 shoot by tying under the uppermost bud or pair of leaves ; if a verti- 

 cally growing leaf is used, the thread may be piit through a tiny hole in 

 its tip. The weight of the pen sliould be so adjusted that the least 

 possible pull will be exerted upon the plant. To set the cylinder 

 truly vertical, or in other words the clock truly horizontal (great care 

 must be taken in removing tlie surplus parts from tlte clock not to bend 

 the works}, use a small thick oiled board (Fig. 24) set level, as tested 

 with a spirit-level, by tapping small wedges under it. The cylinder 

 must be forced down firmly on the spindle or it will wobble. I There 

 there is not room for plant and clock under the same wheel, the former 

 may be set at a little higher level, or else the plant may be set at a dis- 

 tance and the thread led from it to the wheel throitgh a ring of glass 

 tubing or over a needle so arranged that the thread will be vertical 

 from plant to loop or needle and will then reach the wheel in the plane 

 of the latter; but all thread's should be as short as possible in order to 

 avoid hygroscopic changes in length. Half-hour records may be ob- 

 tained from the cylinder by cutting out one-half of the -paper vertically 

 and joining the edges of either piece. It is rather convenient to have 

 a millimeter scale stamped vertically on the paper. 

 The above-described apparatus forms an autographic machine which 

 should be utilizable for other purposes as well. 



Numerous forms of auxanometers have been invented, from the very 

 accurate one of Sachs, familiar in its many modifications, through numer- 

 ous published cuts, down to forms so crude as to be of slight worth. 

 Pfeffer's modification of Baranetzsky's instrument, made by Albrecht of 

 Tubingen, is the best obtainable form. A very exact instrument is de- 

 scribed by Frost in Minnesota Botanical Studies, No. XVII, 1894 ; a crude 

 form is given by Bumpus in Botanical Gazette, 12, 149, and a better one 



