The Climbing Habit in Antirrhinum majus. 289 
twined around the petioles in a direction opposite to that shown in 
Text-fig. 35, A. 
These branches did not develop further, but remained in this 
condition, their growing tips not elongating up to September 28th. 
The majority of the branches, however, had only made one third 
Text-fig. 35. A. Two branches at a node, each coiling around the base 
of its subtending leaf, but one sinistrorse, the other dextrorse. B. Two 
branches at a node, twining in the same direction. The internode above, also 
the upper left-hand branch, have formed sickle-shaped bends. 
to half a turn or in some cases even less. This tendency towards 
the arrest of growth of the branches which developed the twining 
habit was very common but by no means general. On comparing 
Fig. 34, C and E, with Fig. 35, A, it will be noticed that the branches 
in 1908 had formed a coil at the base of the internode, but unlike 
many of those of the present year, this coil did not include within 
it, the stalk of the subtending leaf. On June 14th, the plants were 
carefully examined and an attempt was made to determine the 
length of time occupied by a branch in completing a coil. Twenty 
specimens were marked at points where branches showed signs of 
twining. These were examined at frequent intervals and sketches 
made. Indicators were also attached to show the position of the 
branch at the commencement of the observations. A very consider¬ 
able number of drawings and records were taken, but with somewhat 
disappointing results. In one case, for example, from 7,20 p.m. on 
June 15th to 10.0 a.m. on June 17th, the marked branch had turned 
