My own opinion is that this period is so critical and proper care is so apt _not_ to be taken, that the safest rule is not to begin singing-lessons until the adult voice undisputably has arrived. So many voices have been ruined by lack of care during mutation that it is better no risk should be taken. But why not, it may be asked, have the child taught and, when the period of mutation arrives, have the lessons suspended? There would be no harm in this, excepting that here again is run the risk that proper care will not be taken to stop soon enough and that the career of a possibly fine singer may be ruined. It has happened again and again that voices have been lost irretrievably or impaired permanently by careless use of them during the change from youth to manhood. Therefore, and also because the muscles remain limber and flexible in young people for some years after they have arrived at puberty, I advise that singing-lessons should not begin until the period of mutation is well over. Sir Morell Mackenzie, after stating that the doctrine long has been held universally that not only should systematic training be interrupted, but singing altogether forbidden during that critical period, nevertheless maintained that "_if due care is exercised_ there is no reason why the voice should not be used in singing during the transition period: but the training must be carried out _within certain limits and under strict supervision by a competent person_." But there is so much risk that due care will not be exercised, that those "certain limits" will be overstepped, that the "strict supervision" will be relaxed or not exercised by a "competent person," that I strongly advise not to begin lessons until the period of change is over. In this view I am supported by Garcia, who took sharp issue with Mackenzie. "My father," wrote Garcia, "went through the transition time without ceasing to sing, and without having done himself the least harm. But both my sisters, Mesdames Malibran and Viardot, were obliged to wait a year. I continued to sing, and my voice was ruined!" Continuing, Garcia says that the old rule which has preserved so many voices--that singing should cease altogether during mutation--should not be thrust aside on account of some rare exceptions, and young singers be handed over to the "doubtful caprice of ignorant or careless teachers." A person might with "due care" and "strict supervision" live in a plague-stricken city without contracting the disease, but one would not recommend his going there for his health. Why deliberately expose the voice to danger of loss or permanent impairment by advising that it can be used with safety during the period of transition? Far better to be on the safe side, wait until manhood or womanhood is definitely established, and then begin lessons as soon as possible. CHAPTER III ON BREATHING: INSPIRATION