943 
Oenothera Lamarckiana and O. gigas. 
definition. For these reasons a detailed comparison between the reduction 
processes of Lamar ckiana and gigas is a matter of importance in any attempt 
to establish the fundamental differences between these two forms, and the 
possible manner of origin of the latter type. 
The writer is pleased at this opportunity to describe the conditions in 
gigas , since this plant has been employed as one of the parents in certain 
crosses with wild species of Oenothera . and the behaviour of the chromosomes 
in the hybrids of the first and second generation gives promise of some 
interesting results which, if worked out satisfactorily, will be presented in 
future papers. 
Methods. 
Experiments in methods of fixation have been continued since the 
recent study of Oenothera biennis (Davis, TO), where the effects of Carnoy’s 
alcohol-acetic and chloroform-alcohol-acetic mixtures were contrasted with 
those of chrom-osmo-acetic acid, much to the advantage of the latter fluid. 
In the past summer of 1910 Bouin’s fluid and Juel’s fluid were given 
a similar trial. Anthers of Lamarckiana , after 4-5 hours in Bouin’s fluid, 
were run quickly (1 hour) through grades of alcohol and washed in 80 per 
cent, alcohol until clear. This material exhibited a much greater degree of 
shrinkage than is present in anthers fixed in Flemming’s fluids with the 
precautions outlined in my previous papers. The extent of the shrinkage, 
with no compensatory advantages of importance in the staining reactions, is 
in the writer’s experience unfavourable to Bouin’s fluid as a fixing agent for 
the anthers of Oenothera. 
Juel’s fluid (zinc chloride 2 gr., glacial acetic acid 2 c.c., 50 per cent, 
alcohol 100 c.c.) was tried upon the anthers of gigas , the material being 
left in the fluid 24 hours and then carried through grades of 50-85 per cent, 
alcohol, in which the material was thoroughly washed. The results were 
unsatisfactory. There was great shrinkage of the protoplasts and such 
changes in the structure of the cytoplasm that sharp differentiation of its 
structure was not possible with iron-alum haematoxylin. It was also 
difficult to stain the chromatin and nucleoli, which failed to exhibit th§ sharp 
outlines obtainable after fixation with Flemming’s fluids. 
Further experience with chrom-osmo-acetic acid has not modified the 
methods outlined in the previous papers (Davis, ’09 and TO), but there seems 
to be little choice between the stronger and weaker formulae of Flemming. 
An essential for success with these fluids lies in such precautions as will 
facilitate the most rapid penetration possible, and to this end the writer 
knows of no better practice than that of brushing the anthers with water 
before immersion in the fixing fluid. The greater the attention paid to 
this somewhat laborious technique the less will be the amount of shrinkage 
3 Q * 
