160 Proceedings of the Royal Society 
It is believed that several thousand persons are annually subjected 
to this method of trial, and that the fatal cases are about 50 per 
cent. 
The Akazga arrived in bundles, which consisted of long, slender, 
and crooked stems, having their roots generally attached to them, 
but sometimes their leaf-bearing branches only, and containing 
also a very few complete plants, with roots, stem, and branches. 
The plant is usually about six feet in length ; but some specimens 
were only four, and others as long as eight feet. The bark is yel- 
lowish orange, and in some parts light red, and it is frequently 
covered with a gray efflorescence. It adheres firmly to the stem, 
but may be readily detached, after exposure to a gentle heat for 
some days. Its internal surface is light brown. The space between 
the bark and the wood was found, in a few pieces, to be occupied 
by a large number of minute sparkling crystals ; but it has not yet 
been determined whether these consist of a vegetable or mineral 
substance. The leaves are opposite and oval-accuminate in form ; 
the apex frequently consisting of a linear prolongation more than 
an inch in length. From its general characters, the plant is sup- 
posed to belong to the Loganiaceae, but the materials are insufficient 
to identify it. 
By boiling the powdered bark with alcohol of 85 per cent., 
and distilling and evaporating the tincture, a brown shining ex- 
tract is procured, weighing from 12 to 15 per cent, of the bark em- 
ployed. It has a bitter, non-persistent taste, and, when treated 
with concentrated nitric acid, produces a brownish-yellow colour, 
which is not materially affected by heat, nor by solution of proto- 
chloride of tin. It is obvious that the active principle of Akazga 
is contained in this extract ; and to separate it the following 
method has been adopted, after several attempts with various pro- 
cesses The extract is treated with a very dilute solution of tar- 
taric acid, which removes 77 per cent., and filtered. The clear, 
yellowish-brown acid solution is shaken with successive portions of 
ether, so long as any colour is removed ; and by this means also a 
small quantity of an aromatic oil is separated from it. After de- 
cantation, a solution of carbonate of sodium is added to the liquor, 
so long as it causes a nearly colourless, flocculent precipitate. It 
is again shaken with ether, which is decanted, and agitated with 
