126 THE ROYAL SOCIETY OF CANADA 
Conclusions. 
Summing up our observations and the results of the work in the 
field and laboratory since 1905, we find— 
(1) That as regards the cultures experimented with, while there 
were many instances in which they distinctly favoured the growth of 
the legume, their action on the whole was more or less uncertain. The 
profitable employment of these preparations seems, therefore, prob- 
lematical. We have not attributed the failures always to the culture, 
but we do know that the vitality of these bacterial preparations is very 
quickly impaired by light and heat. Unless made by a reputable firm 
or institution and still fresh, satisfactory results can scarcely be looked 
for. 
(2) The employment, as an inoculating material, of the soil from 
the surface of a field bearing a luxurious crop of the specific legume has 
in our hands given better results than the use of cultures. Where cost 
of transportation is not prohibitive, inoculation with soil will be found 
the most reliable for the general farmer. Provided the proper pre- 
cautions were observed we have never found it to fail on any soil which 
merely lacked the specific nitrogen-fixing bacteria. Notable instances 
of suecessful inoculation by this method have been recorded for alfalfa 
in the North Western provinces. 
(3) Our observations indicate that the nitrogen-fixing bacteria are 
widely distributed over the Dominion. We are of the opinion that in 
the majority of cases, in Eastern Canada at all events, in which there 
has been failure to obtain a good catch of clover, one or more of the 
following causes have been to blame rather than the absence of the 
necessary bacteria: deficiency of moisture; unsuitable mechanical con- 
dition of soil due to lack of humus; inadequate drainage or improper 
working of soil; acidity or sourness of soil, denoting deficiency of lime; 
poor seed; or if the crop failed the second season, the severity of the 
winter. ; 
(4) The systematic analysis of a soil continuously in clover shows 
a constant increase in its nitrogen-content. The experiment reported 
ras continued for nine years and at every examination the soil was 
found to be richer in nitrogen. In spite of losses which must have 
ensued from bacterial activity and other causes, there had been a con- 
stant, though not regular, accumulation of this valuable element. The 
work points to the high manurial value of the residues from a legu- 
minous crop and emphasizes the importance of a rotation which includes 
a legume, if soil fertility is to be economically maintained. 
