standard for a receiving rate is harder and later, and inquiry at
schools where it is taught shows that about seventy-five per cent of
those who begin the study fail to reach this speed and so are not
employed. Bryan and Harter[2] explained the rate of improvement in
both sending and receiving, with results represented for one typical
subject in the curve on the following page.

From the first, sending improves most rapidly and crosses the
dead-line a few months before the receiving rate, which may fall
short. Curves 1 and 2 represent the same student. I have added line 3
to illustrate the three-fourths who fail. Receiving is far less
pleasant than sending, and years of daily practise at ordinary rates
will not bring a man to his maximum rate; he remains on the low
plateau with no progress beyond a certain point. If forced by stress
of work, danger of being dropped, or by will power to make a prolonged
and intense effort, he breaks through his hidebound rate and
permanently attains a faster pace. This is true at each step, and
every advance seems to cost even more intensive effort than the former
one. At length, for those who go on, the rate of receiving, which is a
more complex process, exceeds that of sending; and the curves of the
above figure would cross if prolonged. The expert receives so much
faster than he sends that abbreviated codes are used, and he may take
eighty to eighty-five words a minute on a typewriter in correct form.

[Illustration: Letters per Minute x Weeks of Practice.]

The motor curve seems to asymptotically approach a perhaps
physiological limit, which the receiving curve does not suggest. This
seems a special case of a general though not yet explained law. In
learning a foreign language, speaking is first and easiest, and