Hill . — The Histology of the Sieve-Tit bes of Angio sperms. 259 
it is as well to omit Osmic acid from the killing solutions. The requisite 
amount of swelling, which often need only be very slight, can be brought 
about by a solution of Iodine in an Iodide, or by Picric acid or by Picro- 
Sulphuric acid, for two or three days or longer as circumstances may require, 
and the material may be swollen either in bulk or after it has been 
cut into sections. 
The material may be preserved in Thymol water either before or after 
the swelling of the tissues has been effected, provided the killing, fixing, and 
swelling reagents have been thoroughly washed out. It is also important 
that the material in Thymol water should be kept in glass or india-rubber 
stoppered bottles, as, with ordinary corks, moulds are very liable to appear 
in the bottles. 
Staining. 
The same principles underlie the methods employed in staining as 
have already been published in brief by Gardiner. Certain modifications 
have been introduced to suit special cases, and some new methods have 
been devised, without, however, departing from the general principle which 
appears to govern the method as a whole. 
Staining is effected in all cases after previous mordanting of the proto- 
plasm. If Safranin is to be the stain, then the mordant may be a salt 
of Uranium or Platinum, Potassium permanganate, and also to a certain 
extent Iodine. Uranium and Platinum salts are usually used in conjunction 
with Osmic acid according to a modified formula of either Kolossow or 
Hermann. These latter mordants require careful washing out before the 
staining can take place, and they should be used for pieces of tissue rather 
than for sections. To overcome this difficulty, and also to shorten the time 
taken over the mordanting process, a modification has been introduced 
which is found to be of considerable value in many cases. Since both 
Iodine in Potassium Iodide and Uranium Nitrate are mordants for proto- 
plasm, and the Iodine has rapid penetrating powers, it seemed likely that 
Uranium Iodide, should such a compound exist, would be a very useful 
reagent. Dr. Fenton suggested that if equivalent quantities of solutions of 
Uranium Nitrate and Potassium Iodide were mixed together, when required 
for use, the desired compound might be formed which would mordant the 
protoplasm of the sections placed therein. This has been tried with quite 
successful results ; it is especially useful for dealing with sections, and, should 
the staining be not sufficiently heavy at the first attempt, it is quite easy to 
repeat the whole process one or more times. 
The Safranin method of staining is the most useful for sieve-tube 
research, since it is possible to stain the callus with Water Blue without 
interfering with the stained protoplasm, and it is very important for a right 
understanding of the structure of the sieve-plate that staining of both 
