464 
F. L. PICKETT 
acetic acid i cc, water 99 cc.)- Other fixing agents were used for 
check material but without giving any differences with respect to the 
points under consideration. After the usual infiltration and imbedding 
in paraffin the material was sectioned 5-8 ijl in thickness, and finally 
stained in Haidenhain's iron-alum haematoxylin, in a similar stain in 
which a i/io percent aqueous solution of brazilin replaced the usual 
Fig. 4. A portion of an anther cavity filled with the periplasm. The 
tapetal nuclei have begun to change form and wander among the developing 
spores. X 280. 
}/2 percent solution of haematoxylin, and in various modifications of 
the triple stairk, safranin-gentian violet-orange G. The best differen- 
tiation was obtained by the use of clove oil solutions of both the 
gentian violet and orange G as suggested in the Annals of Botany 
(29: 471-472. 19 1 5). By the use of this stain it was always easy 
to differentiate the tapetal nuclei in the periplasm and to make out the 
smallest details of their structure. The brazilin was found of value 
also in the study of details of nuclear structure. A clove oil solution 
of "Licht-Griin" used after the triple stain above mentioned or after 
the Haidenhain's haematoxylin was found useful in distinguishing 
