IN THE TROPICS. 
359 
afford interesting exceptions to the ordinary routine of Mendelian 
experience. And the work of the practical breeder must usually 
have its beginning in similar speculative experiments ; since from 
the present state of our knowledge no one can foretell with any 
degree of certainty the result of a particular cross which has not 
been previously made. Indeed, one of the chief aims of this class 
of experimental work is the attainment of a system of knowledge 
which will in some degree enable us to foretell the results of new 
combinations ; but this stage of the inquiry is still far from 
being realized. The accurate prediction of phenomena is one of 
the most important effects of all classes of codified scientific know¬ 
ledge, and there can be no doubt that in the case of the present 
branch of physiology a similar degree of cognition is attainable ; 
indeed, the advance in this direction has already been one of 
phenomenal rapidity. 
Theoretical reasons led me to select the following strains of Pisum 
sativum for particular study. The pea Telephone was chosen for 
examination owing to its alleged erratic and exceptional behaviour on 
crossing. Telegraph, a form nearly allied to Telephone, has been 
regarded as showing rather less irregularity. As a fresh specific 
test of these forms they were crossed with a smooth yellow seeded 
pea commonly grown by natives in Ceylon. For studying the colour 
characters of the testa, a large “ French Sugar Pea ” was made use 
of ;* and also a form of Pisum arvense of very primitive appearance, 
and closely similar to one which is known to the natives of North¬ 
ern India as Chota matar, but which is only rarely grown in 
Ceylon. 
From a practical point of view the method adopted was to grow 
a series of good European varieties of peas and to cross those which 
flourished best under the new conditions both with each other and 
with varieties obtained locally. As large a number as possible of 
the offspring of these crosses was then grown, and the progeny which 
showed the best combinations of characters were picked out for 
further propagation. 
Identical, it appears, with the *' Purple Sugar Pea ” described by Bateson. 
