Jan. i3,1923 
New Respirometer for Seeds. 
103 
Other methods of incubation without contact with mercury are, of 
course, possible. But when the mercury is used in contact with the 
respiring seeds, even if no effect of the mercury is known, this possi¬ 
bility should be carefully investigated before depending upon the re¬ 
sults. At best the usefulness of the method is limited by the tedious¬ 
ness of the processes of incubation of the material and of sampling and 
analysis of the gas. 
(4) micro-respirometers, the essential feature of which is the 
USE OF very small quantities of gas or the detection of very 
slight changes in its composition 
Thunberg’s unnecessarily complicated apparatus (27), modeled on 
Peterson’s older C 0 2 apparatus, consists of two connected pipettes of 
similar size—one for analysis of the gas and the other for pressure com¬ 
pensation—each connected with a capillary tube and adjustable bulb of 
mercury. Thunberg’s simpler apparatus, merely for demonstration of 
oxygen consumption (28), was modified by Winterstein (ji), to allow 
the introduction of artificial atmospheres and further modified by Wid- 
mark (30) for accurate quantitative work with very small gaseous ex¬ 
changes. As used by Widmark it is still too cumbersome for general 
use; one half of the apparatus serves merely as a compensator for the 
other half, in which the respiring material is incubated, and the results 
obtained are vitiated by the fact that oxygen consumption and C 0 2 
production are studied in duplicate lots of material instead of in identical 
material, or, if in the same material, then in alternate instead of identical 
periods. Besides, it seems probable that this apparatus could not be 
used when the gaseous exchange is as large as in most studies with moist 
seeds. 
Tashiro ( 26 ) described two apparatuses which he claims will detect 
1 X io” 7 gr. of C 0 2 by absorption in a small drop of barium-hydrate 
solution. The author claims the possibility of accurate estimation of the 
amount of C 0 2 by a series of trials in any given case. These apparatuses 
can not be used with as large quantities of C 0 2 as are usually involved 
in respiration studies without extraordinarily large multiplication of ex¬ 
perimental errors; they are relatively complicated and expensive; they 
take no account of the oxygen consumption; and in spite of the author’s 
claim to the contracry, they seem to be adapted only to rather roughly 
approximate estimation even of the C 0 2 given off. 
Winkler, long ago, described a very sensitive chemical method for 
determining the percentage of oxygen in solutions, using alkaline potas¬ 
sium iodid, manganese chlorid, and hydrochloric acid. Recently, 
Osterhout and Haas ( 20 ) adapted this method for use in biological work 
with aquatic organisms. This method is at best too complicated for use 
when any other method is available. It is not adapted for use with other 
than aquatic organisms, and it neglects determining the C 0 2 produced 
in respiration. 
Warburg (29), by substituting a short column of hot N/80 baryta water 
for the long column of the cold N/10 , was able to determine small amounts 
of respired C 0 2 with a maximum experimental error not exceeding 0.1 
mgm. or 0.05 cc. C 0 2 . 
Later, Krogh (14) developed a manometric apparatus which he de¬ 
scribed as “the simplest form of the closed-space respiration apparatus.” 
Like Thunberg’s apparatus and its various modifications, it makes use of 
