The discovery of the light nuclei
n
0
1
H
1
2
H
1
3
He
2
3
has provided additional and much-needed material on which to base and test any theory of the structure and interaction on nuclear particles. The properties of these nuclei which are best known are their masses, the latest values of which are the following:
n
0
1
1·0083 H
1
2
2·0142 H
1
3
3·0161 He
2
3
3·0172, assuming the validity of the mass scheme proposed independently by Oliphant, Kempton, and Rutherford, and by Bethe (in which the results of the disintegration experiments are found to be consistent if the He
4
:O
16
ration is taken as 4·0034 :16). From these masses the binding energies of the nuclei, considered as combinations of neutrons and protons, may be obtained. We find, then, that H
1
2
= H
1
1
+
n
0
1
- 2·1 x 10
6
e
-volts, H
1
3
= H
1
1
+ 2
n
0
1
- 8·1 x 10
6
e
-volts, He
2
3
= 2H
1
1
+
n
0
1
- 6·9 x 10
6
e
-volts. The most conspicuous feature of these figures is that the binding energy of both H
1
3
and He
2
3
is considerably greater than twice of H
1
2
, and the question arises as to whether this cab be explained without introducing an attractive force between the neutrons in H
1
3
and between the protons in He
2
3
. In this paper we attempt to answer the question by applying the variation method to calculate the binding energies of H
1
3
and He
2
3
, use being made of all available information bearing on the form of the neutron-porton interaction. It is found that definite results cannot be obtained in this direction, but the calculations would seem to indicate that these additional attractive forces must be introduced, and, in any case, upper limits may be found for their magnitude. Before discussing these calculations it is important to examine the velocity of the variation method as applies to H
1
2
by comparing results obtained by its use with exact solutions.
The search for the origin of the light nuclei Li, Be, B
