THE PERCUTANEOUS ABSORPTION OF SALICYLATES 109 • 40 • zo I I I /// ///' /....,,,"•/4 ABRADED ii // • I// // //•/// ,. ' - // /" I I I 0 [ 2 • TIME (HOURS) Fig. 20.--The effect of abrading different amounts of the absorption area on percutaneous absorption of 10% methyl salicylate in mineral oil. absorbed than ethyl salicylate. Isopropyl salicylate is significantly better than all the remaining ones except for N-butyl and phenyl salicylates and after one hour salicyl phosphate, sodium salicylate, morpholine salicylate and diethylamine salicylate produced significantly lower plasma salicylic acid levels than all the others. The plasma ester curves are slightly dif- ferent in appearance perhaps due to differences in distribution, excretion or hydrolysis rate. I I I LIGHT MINERAL OIL METHYL NICOTINATE AND _ -- LIGHT MIN. OIL //.•'/ METHYL NICOTI NATE / /•" POLYETHYLENE _ r /.• -GLYCOL MIX / 'OLYETHYLE.E L"•.,.•*:"-•==":ø-I I I GLYCOL MIX I 2 3 TIME (HOURS) Fig. ?-].--Effect ooe 0.9_.5% methyl nicotinate on percutaneous absorption ooe ]0% methyl salicylate in mineral oil and polyethylene glycol.

110 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Figure 19 demonstrates the fact that acetylation of the phenolic group of ethyl salicylate produced a compound that was much less rapidly absorbed than ethyl salicylate. To measure the effect of mechanical injury upon percutaneous absorp- tion various amounts of the absorbing area were abraded with "Dragon Skin" (a metallic abrading material). The results are diagrammed in Fig. 20. The effect of a rubefacient was tested by mixing 0.25 per cent methyl nicotinate with mineral oil and with polyethylene glycol (50 per cent "Car- bowax 400" -+- 50 per cent "Carbowax 1500"). No significant differences in absorption were produced (Fig. 21). DISCUSSION The results obtained with increasing concentrations of methyl salicylate are in essential agreement with those of Shelly and Melton (12) for the per- cutaneous absorption of histamine. Increased absorption was obtained when higher concentrations were applied in both cases. The apparent con- flict between these data and those of Lang, Vos, Kunze and Urnberger (13) for mercury compounds and of Strakosch and Clark (14) for sulfonamides may be due to the low solubilities of these substances. If the vehicles were already saturated with the compounds at the lowest concentrations used further increases in the amounts of the medicaments would not increase their activities in these vehicles and hence would not increase their rates of absorption. Higuchi (16) in a paper in this symposium has discussed the theoretical considerations of percutaneous absorption. It is likely that the concept of activity of medicament in different vehicles can be fruitfully applied to the results obtained with nonvolatile vehicles. On this basis the signifi- cantly greater rates of absorption of methyl salicylate obtained when pet- rolatum or mineral oil were used as vehicles as compared with lipids and silicone were due to the greater activity of methyl salicylate in these hydro- carbons. Conversely, the significantly lower levels obtained with poly- ethylene glycols was due to a lowered activity of methyl salicylate perhaps due to association with polyoxyethylene groups as suggested by Stolar, Rossi and Bart (10). The unexpected similarity in efficacy obtained with such chemically different substances as silicone, saturated and unsat- urated animal and vegetable lipids is presumably due to the similar activity of methyl salicylate when it is dissolved in these different substances. Halpern, Gaudin and Stiffel (15) have studied the absorption of methyl salicylate, sodium salicylate and diethylamine salicylate by measuring the urinary excretion of salicylic acid. These workers reported that diethyl- amine sa]icylate was absorbed much more rapidly than methyl salicylate and sodium salicylate. Our results lead us to the opposite conclusion. Three preparations containing diethylamine salicylate were tested on rabbit