2 BULLETIN 80, U. S. DEPARTMENT OF AGRICULTURE. 
more steam, and require much larger plants and more labor for the 
same tonnage output per day than others making similar products. 
It was to secure and make available detailed information which would 
both facilitate other experimental work in the laboratory and promote 
the efficiency of commercial plants employing the soda process that 
the series of tests discussed in this bulletin was undertaken. They 
were carried out at the Forest Products Laboratory maintained by 
the Forest Service at Madison, Wis., in cooperation with the Univer- 
sity of Wisconsin. 
The report of the experimental work is prefaced by a short descrip- 
tion of the soda process and a review of previous investigations. 
Some general comments on aspen as a raw material for soda pulp and 
on the pulp itself will be found in the appendix, pages 41-47. This 
species of poplar was selected as the test material because it is the 
most important soda pulp wood. The information secured, however, 
is of much value also in connection with the cooking of other woods. 
THE SODA PROCESS AND ITS APPLICATION. 
What is here referred to as the soda process may be considered as 
a modification of the old Watt and Burgess process, first practiced 
in 1853/ and probably the oldest commercial method for producing 
chemical pulp from wood. It originally consisted in digesting suit- 
ably prepared wood in a large boiler with a strong solution of caustic 
soda under a pressure of about 90 pounds per square inch for 10 or 
12 hours. The wood was then washed to remove the alkali and 
treated with chlorine gas or an oxygenous compound of chlorine. 
The partially digested wood was then washed to free it from the 
hydrochloric acid formed and again treated with a small quantity 
of caustic soda solution. The pulp so produced was then washed, 
bleached, and beaten in a beating engine, after which it was ready 
for the paper machine. The modification of this process as employed 
at the present time in the United States dispenses with the interme- 
diate digestion treatment with chlorine compounds. Different cook- 
ing conditions also are used, the details of which, together with a 
brief description of the manner of preparing the wood, are given 
below. 
PREPARATION OF THE WOOD. 
"While a few mills cook their wood unbarked or only partly barked, 
the general practice is to remove even the live inner bark. 2 The 
i Charles Watt and Hugh Burgess secured a United States patent on this process in 1854. It was devel- 
oped further and modified by Juillon in France (1855), by Houghton in England (1857), and by Albert 
Ungerer, to whom a British patent was issued (1872). Further modifications gradually resulted from its 
commercial application. 
2 The barking loss amounts to about one-fifth of the weight of unbarked logs. The losses in the case of 
logs from 31 trees used in these experiments varied from 16 to 20 per cent, which checks quite well with 
Ziegelmeyer's figure of 19.5 per cent on European aspen. (See Stevens, Paper Mill Chemist, p. 150, 1903.) 
Aside from the convenience and ease of barking in the woods, the saving of freight is considerable when the 
wood is transported to the mills by railroad, and since the barked wood dries out rapidly an additional 
advantage is secured by the loss of weight in seasoning. A cord ol green aspen (about 50 per cent water) 
weighs about 1.900 pounds more than the same wood, air dry (about 15 per cent water). 
