1386 MESSRS. J. B. LA WES, J. H. GILBERT, AND M. T. MASTERS, 
plants. These characters were found to be quite consistent with some main differences 
in the chemical composition of the produce. 
Thus, on plot 10, where the supply of potass was discontinued, the tendency to 
form stem has gradually, and at length very greatly, diminished; the herbage has come 
to consist in a much larger proportion of leaf; it maintains a darker green colour, 
does not eventually ripen as on plot 9, and is altogether more patchy and uneven. 
The chemical examination of the produce shows that, with these characters, sub¬ 
stantially the same amount of nitrogen continues to be taken up, on plot 10 with the 
discontinued supply of potass and the diminished and deficiently matured produce, as 
on plot 9. On the other hand, the amount of potass taken up has, in the later years, 
been only about one-third as much on plot 10 as on plot 9. We have here again, 
then, so far as the nitrogen taken up is concerned, and also the abundance of 
chlorophyll formed, as indicated by the colour of the herbage, in a sense, potential 
growth. But, with the deficient supply of potass, there is deficient stem-formation 
and immaturity ; in fact, deficient produce ; in other words, deficient carbon-assimila¬ 
tion. Nor did the increased supply of soda compensate for the deficiency of potass, 
so far as these characters of growth are concerned. 
From the various facts adduced, it is evident that it was the want of a liberal 
available supply of potass that restricted the growth on plot 4-2, with the ammonia- 
salts and superphosphate of lime, and on plot 10 with the ammonia-salts, super¬ 
phosphate of lime, and a liberal supply of soda and magnesia, but only a partial one 
of potass ; indeed, the agricultural, the botanical, and so far the chemical results, 
concur in showing how essential is a liberal provision of potass for the fu ll effect of 
the nitrogen which was at the same time supplied. The comparison of the results 
obtained on plots 7 and 8, the one with the mixed mineral manure each year, including 
potass, and the other with the potass discontinued after the first few years (as on 
plot 10), but in both these cases without ammonia or other artificial nitrogenous 
supply, showed that, under those conditions also, the potass was the most essential 
to the growth among the mineral constituents supplied. But, whilst when the potass 
was used in conjunction with ammonia, its effect was greatly to develop the Graminese, 
and practically to exclude the Leguminosse, it was, when used without ammonia, very 
characteristically to increase the Leguminosse, and very much less the Graminese than 
when the ammonia was also applied. 
14. Ammonia-salts alone (400 lbs. per acre), 13 years (1856-1868); succeeded by 
mixed mineral manure, including potass, each year since , Plot 6. 
From the commencement of the experiments, plots 5 and 6 each received the same 
amounts of ammonia-salts without any mineral manure, and the plots were, therefore, 
so far, duplicates. The only difference was, that during each of the first seven years 
plot 6 received sawdust in addition, which, however, had little or no effect. 
