Commencing with a one grain dose at about 6:30 P.M., on alternate evenings, leaving the intermediate evenings in order to be sure that the nightly tendency still persisted, and increasing by half a grain each alternate evening, no very definite effect was perceived, until the dose reached 21/2 grains, and this dose simply rendered the tendency to sleep resistible by effort.  After an interval of three evenings, with the tendency to sleep recurring with somewhat varying force each evening, a dose of 3 grains was taken, the maximum single dose of the German Pharmacopoeia.  This gave a comfortable evening of restedness, without sleep or any very strong tendency to it until ten o’clock.  Without anything to counteract sleep, the rule was to read with difficulty by nine, without much comprehension by quarter past nine, and either be asleep or go to bed by half past nine.  The 3 grain dose of caffeine repeatedly obviated all this discomfort up to ten o’clock, but did not prevent the habitual, prompt, and sound sleep, from the time of going to bed till morning.

This was the model established, upon and by which to measure all the other agents, and they were never taken nearer than on alternate evenings, with occasional longer intervals, especially when the final doses of record were to be taken.

The next agent tried in precisely this same way was coca; and knowing that the quality of that which was attainable was very low, the commencing dose of the leaf in substance was 2 drachms, or about 8 grammes.  This gave no very definite effect, but 21/2 drachms did give a definite effect, and a subsequent dose of 21/2 fluid drachms of a well made fluid extract of coca gave about the same effect as 21/2 grains of caffeine.  Three fluid drachms of the fluid extract were about equivalent to 3 grains of caffeine.

Both the coca used and the fluid extract were then assayed by the modern methods, for the proportion of the alkaloid they contained.

The only assays of coca that could be found conveniently were those of Dr. Albert Niemann, of Goslar, given in the American Journal of Pharmacy, vol. xxxiii., p. 222, who obtained 0.25 per cent.; and of Prof.  Jno.  M. Maisch, in the same volume of the same journal, p. 496, who obtained 4 grains of alkaloid from 1,500 grains of coca, which is also about a quarter of one per cent.  These assays were, however, very old, and made by the old process.  The process used by the writer was the more modern one of Dragendorff slightly modified.  It was as follows: 

Thirty grammes of powdered coca, thoroughly mixed in a mortar with 8 grammes of caustic magnesia, were stirred into 200 c.c. of boiling water, and the mixture boiled for ten minutes.  The liquid was filtered off, and the residue percolated with about 60 c.c. of water.  It was then again stirred into 150 c.c. of boiling water, and was again boiled and percolated until apparently thoroughly exhausted.  The total