758 H. E. KNow.LTon 
It is evident, from a study of these experiments, that the pollen of. most 
species remains viable longest under conditions of low temperature. 
Altho there are few available data, the indications are that the optimum 
moisture conditions vary with the species. 
In all of the experiments mentioned in the foregoing paragraphs, arti- 
ficial germination, except as otherwise noted, was the only test of viability 
used in the investigations. As will be shown by the author’s experiments, 
this is not always a proof of ability to bring about fertilization. 
Possible causes of death of pollen 
As far as the writer has learned, Andronescu (1915) is the only investi- 
gator who has advanced a theory to explain the cause of death of pollen. 
He found that corn pollen stored out of doors lost 48 per cent of its moisture 
in two hours and 52 per cent in twenty-four hours. Since he found also 
that pollen lived longer at higher humidities, he concluded that death is 
the result of desiccation. Pfundt (1910), however, determined that the 
range of moisture required is wide, depending on the species: some live 
longer at low, others at high humidities. It is not possible to conclude, 
‘therefore, that desiccation is always the cause of death. 
Investigators generally agree that the duration of life is longer at fairly 
low temperatures. This is to be expected, as physiological activities are 
slower. 
Altho pollen and seeds differ morphologically as well as in function, 
it was thought that the cause of death in each might be similar. A brief 
discussion of the causes of death in seeds is, therefore, desirable.  , 
Acton (1893) found that thirty-years-old wheat which had lost its germi- 
native power, contained about the same amount of stored food but less 
water than did new grains, and that its amylase and proteolytic enzymes 
had also been destroyed. On the contrary, Brocq-Rousseu and Gain 
(1909) state that amylase and oxidase were present in wheat grains . 
ranging from twenty-five to eighty years old. White (1909) states that 
amylase was present in fairly large quantities in old seed of wheat, barley, 
oats, rye, and corn. The age of the seeds tested ranged from two to 
twenty-one years. The addition of amylase did not cause dead seeds 
to germinate. 
Crocker and Harrington (1918) and their coworkers found that, in 
some seeds, catalase activity is correlated with physiological activity and 
