(155 ) 
may be well again to emphasize the fact that there is no doubt 
of this wide range of movement, because the scar of the re- 
moved leaflet in all cases served as a point of orientation in 
determining the identity of the leaflets. 
Before going on with the analysis of Table VI let us take 
the averages as they stand. Considering the great range of 
positions which the operated series includes the differences 
of the averages show a striking similarity to those of Par- 
thenocissus. In fact leaflets C and B show almost identical 
movements in the two cases. In A and & however the con- 
Fie. 8 (X1%.) Lupinus albus (all cases), Average normal (un- 
broken lines) and average operated (dotted lines) position and length of 
leaflets. (The dotted line at # is incorrectly placed ; it should be 11.3° from 
the unbroken line. ) 
siderable number of cases in which the leaflets have moved 
over the surface of the petiole has left a minus influence in A 
and a plus influence in £ (see f. 8), but in neither A nor & 
is the change sufficient to destroy the general similarity to 
Parthenoctssus and we have here the same tendency toward 
the production of a symmetrical four-leaved system. 
