The uncertain rate of progress in teaching drawing, the want of fixed principles that has prevailed in the instruction of it up to this day, even from remote ages, have long since rendered it desirable to have a method surer in its results, and capable of being applied by all teachers alike.

Anything like demonstration is impossible by the ordinary methods of instruction: the different ways in which the masters may regard the instruction and the art itself become the rule – a very variable one, as we can imagine – that governs the schools. Even admitting that these different roads can lead to an almost common result, that is to say, to a satisfactory knowledge of drawing, it is easy to see how important the functions of the master become, and how necessary it is that his special talents should qualify him for guiding the pupils in the midst of the uncertainty of the rules.

The first difficulty in such a method of instruction consists, then, in finding a sufficiently large number of teachers endowed with indispensable talents, and resigned to the exercise of functions that are, of course, poorly recompensed.

The second, and perhaps the most insurmountable difficulty, consists in the impossibility of procuring good models. Those that are met with in the schools, produced in all the successive styles, chosen hap-hazard, devoid of correctness or expression, can only vitiate the pupil's taste, and render the best guidance almost useless.