July, 1923 ] 
SANDS — STRUCTURE OF CHROMOSOMES 
345 
The structures here are more or less obscured and are, in fact, rarely 
seen in the material prepared by the standard killing solutions and imbed¬ 
ding methods. A comparison of Tradescantia prepared by these two 
methods convinces me that, in the sectioned material, the finer details of 
the living cells are lost by slight fusions, shrinkage, and distortions, which 
leave perhaps the relative proportions as a whole unimpaired. 
By dilution technique, any desired degree of staining intensity may be 
obtained ranging from none or the natural hyaline to that of deep color. 
As stated, the most frequently used dilution was, perhaps, one drop of 
aceto-carmine to one drop of water. The rate of disintegration suffered 
by the mounts prepared with this dilution was usually so rapid that one 
could not study the preparation for several consecutive days. Weaker 
mixtures were therefore devised. 
Dilutions were made in a series of test tubes, so that tube A contained 
a one-to-one mixture; tube B, a one-in-four; tube C, a one-in-eight; tube D, 
a one-in-sixteen; and tube E, a one-in-thirty-two. If the tubes are cleaned 
of all foreign inorganic and organic matter, the stain will not precipitate 
and the tubes may be kept corked. Paraffined corks should be prepared, 
otherwise evaporation may be considerable. For most anthers, tube E 
contained too much water in proportion to acetic acid so that the effects 
of toxic stimulus generally appeared. 
This method was found preferable to one in which pipettes graduated 
to hundredths of a cubic centimeter were employed for making the dilutions 
directly on the slide. Too much inaccuracy resulted from the latter practice. 
The 1:16 dilution was very satisfactory, and permitted one to observe 
the cells unstained for a long period of time, then through all intermediate 
stages of slow color absorption till at last the chromatin contents are of 
a deep rose color. This latter may require anywhere from six to twenty 
hours, since the contents of the anther sacs do not all react alike to the 
staining fluid. Some take up the stain more slowly than others. I found 
that with weaker dilutions the cells would often burst; in order to overcome 
this the dilutions were made more nearly isotonic by the addition of cane- 
sugar solution instead of water. 
The dividing pollen mother cells of Tradescantia virginica L. treated 
as above described show clearly that the chromatic elements are composed 
of bodies roughly similar to those described by Balbiani (1881) and by 
Pfitzner (1882). The shape and size of these granules and their arrange¬ 
ment or distribution in the chromosome have not, however, been ade¬ 
quately described. 
In figures 11 and 13, Plate XXIX, some of the strands are seen to be 
apparently made up of two rows of granules side by side. By numerous 
authors, these granules have been spoken of as chromomeres , a term intro¬ 
duced by Fol (1891). Eisen (1900) distinguishes chromioles as components 
of chromomeres. I also consider the bodies making up the chromosome 
