1897 | RELATION OF NUTRIENT SALTS TO TURGOR 401 
the larger size seemed unnecessary, and was much less conven- 
ient. The seeds were germinated in clean sawdust, and when 
the radicles were a few centimeters in length, average specimens 
were selected from a large number, placed in the culture jars, 
and plugged fast with cotton. When the duration of the experi- 
ment demanded it, the solutions were renewed, but in many 
cases it sufficed to keep the jars full by occasional addition of 
distilled water. 
The first solutions were prepared according to de Vries’ 
isotonic coefficients 3: afterward they were made according to 
molecular equivalents. The latter is the more scientific plan, 
but does not differ enough to affect the results. All salts used 
were of absolute purity, and the utmost care was taken that the 
concentration was exactly that stated: for instance, in making 
up MgSO, solution all dehydrated crystals were removed indi- 
vidually before the salt was weighed; and such compounds as 
Ca(NO,),, which cannot well be weighed dry, were made in 
solution synthetically by the writer. Clarke’s atomic weight 
determinations were used. Making up the solutions by exact 
molecular equivalents secured as nearly as was possible their 
osmotic equality. For the individual salts, in solution alone, this 
might have been accomplished more accurately by taking into 
account their varying degrees of dissociation. But as the oppor- 
tunity for unavoidable and unseen breaking up and reformation 
of molecules in such dilute solutions is limited only by the 
variety of salts dissolved, there is no way of determining the 
Salts whose dissociation should be reckoned with. At any rate, 
in solutions so dilute as those used dissociation of all nutrient 
salts is very complete, and in practice may be regarded as quite 
so. 
The turgor was determined by the usual method, 7. ¢., by 
immersing sections in a sequence of solutions of known strength 
of KNO,. In the experience of the writer, no other solvent 
_has proved itself so available for this purpose as saltpeter. It 
is easily and accurately prepared, keeps well, diffuses quickly 
*3 Op. cit. 
