238 MEMORIES OF MY LIFE
experiment. Sir Andrew Noble kindly undertook to make experiments for the Office, using a i o-pounder gun that happened to be at the Armstrong Works at Elswick. It had been designed especially for shooting at balloons, and was furnished with the necessary spring for preventing harm from recoil. The results were very good and consistent. The shells burst at a constant height of about 9000 feet, and gave a conspicuous and durable cloud of smoke, whose drift could be easily seen and its rate calculated. I designed a camera-obscura arrangement to do this conveniently. The recorded interval of time between the explosion as seen and as heard, was an adequate measure of the distance of the shell-burst. I t could be ascertained with more care when desired, and in more than one way. The cost of each shot was about ten shillings. This method of observation was not followed up, as none of the existing stations were thought suitable, and it was difficult to find one that would be so, considering that easy telegraphic connection with the Meteorological Office was a necessity. Again, the method would be useless in cloudy weather. It may possibly be of future service for inquiries into the varying thickness of the Trade winds in particular localities.
Yet another attempt of mine may be mentioned. Chiefly through the initiative of Admiral Fitzroy, " Wind roses," as they are called, were calculated for the various Ocean districts, bounded by lines of latitude and longitude io degrees apart. They formed adjacent rectangles or " squares" in the maps used by seamen, which are always drawn on " Mercator's projection." The " rose " consists of