But entirely apart from such serious intervals of mal-employment as this, is it not probable that all of us are mal-employed for some little part of our time? Is it not probable, in other words, that our work would be improved if we should omit certain parts of it and do nothing at all instead? It is certain, for one thing, that no one could work continuously, day and night, without serious or fatal mal-employment. That is the reason why our working hours are limited to seven or eight in the twenty-four. Doubtless some workers are over worked and thus mal-employed in their hours of overwork--the sleepy railroad engineer, for instance, who misses a signal and sends a hundred passengers to eternity. We are doubtless free in the library from just this kind of mal-employment, except so far as it is forced upon us by assistants who work or play too strenuously outside of working hours. To go back to the assistant who is cross or careless for an hour every day; it is quite possible that she is in no condition for working during that hour; and this is not because the library hours of work are too long, but because she does not take needed rest outside of those hours. Sometimes this cannot be helped; often it is distinctly the worker’s fault, and it is surely putting the library in a false position to make it overwork its staff to their detriment and its own, just because the assistant puts in her best and freshest hours in work, or more often in amusement, outside the library. Let me pause here to say that the reason we take vacations is to avoid the chance of this kind of mal-employment. The theory of the vacation is widely misunderstood. Some take it to be a period of amusement granted for services rendered. “I think I have earned a vacation,” they say. Others look upon it as play-time wrung from an unwilling employer--the more they can get the better off they are. Few realize that it is, or ought to be, simply an incident in the year’s work, an assignment to special duty, without which mal-employment would be more apt to result.