172 BOTANICAL GAZETTE [SEPTEMBER 
In all cases the dried seeds remained for many days (16-30) in the 
fluids without showing any injury, as shown by their normal 
germination afterward. That these fluids do not penetrate the 
testa can be demonstrated indirectly by using seeds with defective 
coats. Inno instance were such seeds found to be viable after more 
than about 15 hours’ immersion in absolute alcohol, ether, chloro- 
form, or acetone. The length of time seeds with broken coats can 
remain alive in these substances depends upon the size and position 
of the defect. If the break occurs immediately over the hypocotyl, 
a few hours’ immersion will kill the seeds; but if at the tips of the 
cotyledons, they will remain viable some hours longer. 
The question naturally arises as to how much moisture can be 
put into these liquids, and still leave the seeds uninjured by long 
soaking. Commercial 95 per cent alcohol was found not to kill 
merely air-dried seeds in 4 days’ time, but had killed them within 
8 days. A series was then run in go, 80, 70, 50, and 35 per cent 
alcohol. In a few hours the seeds in the lowest three grades had 
become excessively swollen and semitranslucent; they were found 
to be dead. Those in 80 per cent were by no means so swollen, but 
had been killed. Those in 90 per cent were apparently hard and 
sound, but at the end of 3 days all were dead. They may have been 
killed in considerably shorter time than this, as they were not tested 
for germination till the end of 3 days. 
The results of these experiments confirm the findings of BEc- 
QUEREL, who kept various seeds in alcohol, ether, and chloroform 
for a full year without any serious loss of viability. While these 
facts seem remarkable, they merely confirm the results reported a 
good many years ago by GIGLIOLI (21), who obtained, for instance, 
the germination of 20 per cent of alfalfa seeds after more than 
16 years in a saturated solution of HgCl, in absolute alcohol. 
GIGLIOLI says that many of these seeds produced plants which 
flowered and fruited normally after this long exposure to destructive 
gases and liquids. The protective nature of the dry seed coats of 
certain seeds, therefore, is fully established, and is due to imper- 
meability. : 
The results of these experiments offer no evidence against 
BECQUEREL’s view that gases, especially oxygen, are not able to 

