( 27 ) 
THE OVULE OF THE BE UNI AC E^. 
By W. T. Saxton, M.A., RL.S. 
(Communicated by L. Peringuey, F.R.S.S.Af.) 
(Read March 16, 1910.) 
Introduction. 
In a recent essay by Schonland (3) the BrmiiacecB are mentioned as 
one of the more ancient constituents of the South African Mora. The 
author remarks: ''The BrimiacecB . . . are so isolated amongst Saxi- 
fragineae that we can also only point to their antiquity." It seemed, 
therefore, desirable that an investigation of the structure of the ovule and 
embryo-sac should be undertaken. 
The family is endemic in South Africa, and out of the 12 genera 
4 are represented in the Cape Peninsula, i.e., Brimia, Berzelia, Staavia, 
Audouinia. Fortunately, these well illustrate the range of types met 
with in the order, and it is very unlikely that any other genus would 
differ widely from these four. One species of each genus has been 
investigated, and the structure has been found to be very constant 
throughout. The primary object of the research was the investigation 
of the structure of the mature ovule, but it was necessary to study 
certain phases of development in order to explain the mature structure, 
and these are therefore included in the present account. The methods 
used have been those usually employed in cytological work, the fixing 
agents used having been (i.) chromacetic acid — 1*5 grams chromic oxide, 
7 c.c. acetic acid, 250 c.c. water ; (ii.) picric, acetic, corrosive sublimate 
solution. The latter was found the most satisfactory. 
An exceedingly accurate account of the position and external 
characters of the ovule in the genera investigated is given by Bentham 
and Hooker (1). In Berzelia the ovary is unilocular and there is only 
a single pendulous, anatropous ovule. In Staavia and Brunia the ovary 
is bilocular, in the former genus with one ovule in each loculus, in Brunia 
iwdiflora with two. In Audouinia the ovary is trilocular with two ovules 
