By Admiral Fitz-Roy, F.R.S.

As an excuse for writing rather positively about weather and its premonitory indications, I beg to say that nearly half a century ago I was taught by my father, a farming, gardening, and fox-hunting country gentleman, the use of barometers, thermometers, and hygrometers, as "weather glasses;" and it is more than 40 years since I began to use such instruments regularly in the South Pacific Ocean.

" Weather glasses " were used in the 17th century. Among others, De Foe watched and registered them in 1703 (see " The Storm") ; but it is an instance of the necessity for repeating information, that, generally speaking, even now little real use is made of these instruments, however familiar, common, and inexpensive they have become.

Like seamanship, ability to foretell weather is acquired, by degrees, practically, and has not been hitherto attained by books; but it may now be so, in consequence of numerous recorded observations and opinions, brought together in late years and very carefully considered.

Instructions are now available, based on scientific as well as practical conclusions, with which help, properly studied, any one may become " weather-wise" who will notice, even once a-day, the indications of the heavens, of thermometers, and of a barometer.

Marked distinction is advisable between such observations and instructions as are intended only for indicating changes of weather, or its duration, and those of a superior kind required for comparisons and elaborate deductions for scientific purposes.

Want of such sufficient distinction has been hitherto prejudicial.

Unduly mixing scientific objects with mere practical notices of weather has had a tendency to repel the less educated, or to accumulate inferior records unavailable for really scientific purposes.

To require that a perfect barometer, equal to the demands of modern science in an observatory, should be used as a mere " weather glass" at a life-boat station or fishing village, might remind one of putting a racehorse in a cart, or using a razor to cut sticks.

As discussion has arisen about " boiling mercury" and " inspecting the process/' I think it a duty to endeavour to throw some light on this subject.

To know whether a tube with mercury has been well boiled, as it is called (by holding and turning it over a charcoal fire), it is unnecessary to watch the tedious process, as examination of the metal in the tube, with a lens, and its " click" at the top ( the tube, give unfailing evidences of the presence or absence of air, whether boiled or otherwise treated.

To verify the graduation thoroughly (not at a few casual heights only, by comparison with another barometer), artificial pressure or exhaustion must be obtained by placing the instrument under the receiver of an airpump.

This is done at Kew, very completely: and it is necessary for accurate scientific barometers, but not for mere weather glasses.