1921] 
DUGGAR & KARRER—SIZES OF MOSAIC DISEASE PARTICLES ' 353 
of course, to dialyze or otherwise further purify the product, as 
was requisite in the type of studies pursued by Bottazzi (713), 
Reichert (709), and others. 
For ultrafiltration work hemoglobin has been recognized as a 
product of exceptional value. By any standard method of 
preparation it would seem that the particles are of fairly uniform 
size, so much so that it was employed by Bechhold in standardiz- 
ing and designating the porosity of his gelatin filters. Neverthe- 
less, the actual sizes of the particles do not seem to have been 
determined. In one of his papers Bechhold (’07) indicates that 
the particles must average a little less than 20 pu, being fairly 
comparable with ‘‘Kollargol (koll. Silber v. Heyden)”. In his 
text (19), moreover, the same author places them at smaller 
than the particles of 1 per cent gelatin and larger than serum 
albumen, which would indieate a measurement somewhat 
greater than 30 yy. Later in the same work (p. 111) he indicates 
the sizes of hemoglobin particles at 33-36 uu. The diameter 
of the hemoglobin molecule has been given as 2.3-2.5 pu. 
The tests with the standardized hemoglobin solution yielded 
results both satisfactory and illuminating. "Through the Mandler 
filter with the usual time interval mentioned the filtrate was a 
very deep red, yielding no appreciable dilution of the hemoglobin. 
Through the spherical atmometer cup the filtrate was still very 
red, indicating that relatively few particles of the hemoglobin 
were held back. Through the cylindrical atmometer cup A 
there was a very slight passage of hemoglobin particles, while 
through the cylindrical tube B and both tubes impregnated with 
Al(OH), there was no passage of hemoglobin particles whatsoever 
in these tests. 
Further it may be of interest to state that the spherical at- 
mometer cup referred to above permitted approximately 50 
per cent of the gelatin particles to pass fhrough the filter from 
a 1 per cent solution of gelatin. The amount passing through 
was determined colorimetrically in comparison with the 
original solution by means of the Biuret test. The gelatin 
solution was prepared by adding gelatin to the boiling water 
and then immediately cooling to room temperature. 
