The books one reads abound in erroneous dates, mistaken names, garbled extracts, and blundering quotations. So much the more important is it to the librarian, who is so continually drawn upon for correct information upon every subject, to make sure of his facts, before communicating them. When (as frequently happens) he has no way of verifying them, he should report them, not as his own conclusions, but on the authority of the book or periodical where found. This will relieve him of all responsibility, if they turn out to be erroneous. Whenever I find a wrong date or name in a printed book, or an erroneous reference in the index, or a mis-spelled word, I always pencil the correct date, or name, or page of reference in the margin. This I do as a matter of instinct, as well as of duty, for the benefit of future inquirers, so that they may not be misled. I speak here of errors which are palpable, or of the inaccuracy of which I have positive knowledge; if in doubt, I either let the matter go entirely, or write a query in pencil at the place, with the presumed correct substitute appended. Never be too sure of what you find in books; but prove all things and hold fast to those only which you find to be beyond dispute. Thus will you save yourself from falling into many errors, and from recanting many opinions. It is the method of ordinary education to take everything for granted; it is the method of science to take nothing for granted. I may refer here to another rule always to be observed, and pertaining to the theme of strict accuracy in your daily work. That is, the necessity of carefully examining every piece of work you may have done, before it leaves your hands, for the purpose of correcting errors. All of us are not only liable to make mistakes, but all of us do make them; and if any one has a conceit of his own accuracy, the surest way to take it out of him is to let him serve an apprenticeship in some library, where there is competent revision of all the labor performed. There are multitudes of assistants in libraries who cannot write a letter, even, without making one or more errors. How often do you leave out a word in your writing experience, which may change the meaning of a whole sentence? So, in writing titles, whether for the catalogue, or for a library order, or for the information of some inquirer, you are liable to make errors of date, or edition, or place of publication, or size, or to misplace or omit or substitute some word in the description of the book. There is nothing in the world quite so easy as to be mistaken: and the only remedy (and it is an all-essential one) is to go over every line and every word of what you have written, before it leaves your hands. As second thoughts are proverbially best, so a second careful glance over a piece of writing will almost always reveal some error or omission to be corrected. Think of the mortification you must feel at finding an unverified piece of work