214 
pulled on this cogwheel by means of an armature with a pawl. A 
spring drew back this armature just as far that, with the next 
closing of the circuit, the cogwheel should go on one, two or more 
teeth. This simple method, however, was not reliable, as the wheel 
sometimes turned too far, after the shocklike movement. 
This error could only be eliminated by a rather complicated 
mechanism, ingeniously devised by Mr. pr Bourer. Each time, when 
a current passes through the coil of electro-magnet (M!) the armature 
(AI) is attracted. This armature turns around an axis (18). (See fig. 2). 
On the armature is fixed a lever with a tooth (19), pressed against 
cogwheel (20) by means of spring (21). On the axis of cogwheel 
(20) is fixed a spiral-spring (22) which is wound up, when the 
armature (Al) is attracted. This spring tries to relax itself on a 
little clockwork, consisting of some toothed wheels (k,, rj, k, and r,). 
On the axis (28) of toothed wheel (k,), however, is fixed an escapement- 
wheel (24) with a single tooth (25), which is. held, up by an 
escapement-lever (26). On the same axis (28) is fixed a little bolt 
(27) which, when revolving, should catch a tooth of cogwheel (17) 
of the auxanometer. 
The escapement-lever (26) is one with the lever (28) that prevents 
the direct relaxing of spring (22). When the armature (A!) is 
attracted and spring (22) is wound up, the lever (28) makes way, 
slipping over a tooth of cogwheel (20). In the mean time, the 
escapement-lever (26) makes way and relaxes the tooth of escapement- 
wheel (24), at the moment when the lever (28) slips over the top 
of the tooth. The escapement-wheel (24) makes one revolution, bolt 
(27) too, implicating in its revolution cogwheel (17) over a certain 
distance. By means of changing the length of bolt (27), one can 
adjust very accurately the number of teeth that cogwheel (17) shall 
turn. Two teeth (= 10 u) proved to be the most practical arrangement. 
The gradual movement of the cogwheel (17) warrants a high 
accuracy. The proportions of the teeth-number on the wheels (4) 
and (r,) is chosen in such a way, that the elockwork is relaxed 
exactly as much as it is wound by the attraction of the armature 
(Al) .(4,:7, = 8:1). The relaxation-velocity of the clockwork is 
moderated by a fan (29). 
The anxanometer described is capable of recording a total increment 
of the test-plant of 3,5 c.M. This is a considerable amount for 
seedlings. After each experiment, the contact-device is put down 
into its lowest position by turning the nut (m). The machine has 
been tested by the micrometer-screw of a Zeiss I A microscope. It 
has been placed in a room, for constant temperature and shortly 
