COMMON EARTHENWARE JARS A SOURCE OF ERROR 
IN POT EXPERIMENTS 1 
By J. S. McHargue 
Research Chemist , Department of Chemistry , Kentucky Agricultural Experiment Station 
In an investigation to determine whether or not manganese is neces¬ 
sary for the normal growth of plants, by means of carefully prepared pot 
cultures, occasional results were obtained in the control pots which 
indicated that the plants were obtaining manganese from an unrecog¬ 
nized source. Since manganese had been carefully eliminated from the 
sand and the mineral nutrients mixed with it, it was evident that the 
pot was the source of manganese, although the pot was clean and appar¬ 
ently well glazed on the inside surface at the time the nutrients were 
added. 
It had been observed previously that among the 80 pots in use in this 
experiment there were a few on which crystalline deposits of mineral 
nutrients appeared on the outside after they had been wet a few times. 
This fact showed that the walls of the pots were porous and not suffi¬ 
ciently well glazed to prevent the migration of moisture which carried 
the mineral nutrients in solution through the walls so that subsequent 
evaporation and deposition of the mineral nutrients occurred on the 
outside. Judging from external appearances, these pots were as well 
glazed as other pots on the outside of which no deposit of mineral nutri¬ 
ents occurred. 
In Plate i, the only plate accompanying this article, and in references 
to which only the letters A, B, and C will be used, A shows the extent 
of the migration and deposition of the mineral nutrients through the 
walls of the pot. The white, frosted material which appears plainly on 
the brown glaze extended practically over the outside surface of the pot. 
Pots similar in grade to those shown in A and C are in common use in 
pot experiments at agricultural experiment stations. 
The observation that a few of the total number of pots were sufficiently 
porous to allow mineral nutrients to migrate through their walls suggested 
the idea that other similar pots might have walls sufficiently porous to 
absorb, from soils or sand used in culture experiments conducted in 
them, nutrients which would affect the results of other experiments 
made in the same pots at a later time. 
This conjecture is supported by results obtained in experiments with 
manganese. Tomato plants were grown in pots that had been previ¬ 
ously used in other experiments and were similar in grade to the pot 
shown in A. No deposit of mineral nutrients occurred on the exterior 
of any of these pots when like amounts and kinds of mineral nutrients 
were mixed with the sand in the several pots. 
C represents two of these pots containing tomato plants that were 
grown in purified sand and mineral nutrients. Manganese was carefully 
excluded from the sand culture on the left, whereas the one on the right 
contained 0.25 per cent of manganese in the form of the carbonate. 
The sand cultures were kept at the proper moisture content by frequent 
1 Accepted for publication July n, 1923. 
(*3 I ) 
Journal of Agricultural Research, 
Washington, D. C. 
ahd 
Vol. XXVI, No. 3 
Nov. 3, 1923 
Key No. Ky.-i4 
