1895. ] The Embryo-Sac of Aster. 211 
Other Composite. 
In pursuing these studies, I have made preparations of sev- 
eral other Composite, and in Solidago especially I have found 
considerable irregularity in the number and arrangement of 
the antipodal cells, but in no other have I found such exten- 
sive variation as in Aster Nove-Anglia. 
Methods. 
Absolute zlcohol, saturated aqueous solution of picric acid, 
and I per cent. chromic acid were used in fixing material for this 
work. The smaller heads were merely halved; the larger 
heads were cut into sections about one-eighth of an inch 
thick before placing in the fixing fluid. Picric acid gave as 
good results as the chromic. With either acid the fixing is 
complete in twenty-four hours. Picric material should be 
washed in 70 per cent. alcohol until all yellow color disap- 
pears. Chromic material should be washed at least twenty- 
tour hours in cold water or twelve hours in warm. In either 
case the water should be changed frequently. 
Xylol, followed by a mixture of xylol and paraffine was 
used to precede the paraffine bath. Rosen’s method was also 
quite satisfactory. After dehydration it is briefly this: (a) 
equal parts of absolute alcohol and bergamot oil; (6) pure 
bergamot oil; (c) equal parts of bergamot oil and paraffine at 
40° C; (d) soft paraffine; (e) hard paraffine. 
All sections were serial. I did not find it necessary to cut 
thinner than sy. 
With the exception of a little material which was stained 
in bulk with alum carmine, all sections were stained on the 
slide. Delafield’s haematoxylin, acid fuchsin, Bismarck 
brown and Biondi-Ehrlich are excellent stains for this work. 
After many slides have been rinsed in the distilled water jar, 
the water becomes deeply colored. Some slides left over 
night in this jar to await staining in the morning showed a 
striking differentiation. My most satisfactory staining was 
subsequently obtained in this way. 
Summary. 
I. (a) The early development of the macrospore of Aster 
Nove-Anglia differs little from described types. 
(4) The formation of the secondary nucleus has no rela- 
tion to a sexual process. 
