When built all proved able to steam very well, but after some years service certain of them fell off very badly in speed. Others, however, remained as fast as when they were built--the _Terpsichore_, in 1908, averaging 20.1 knots, and the _Aeolus_, in 1909, nearly nineteen knots. During their service, the _Melampus_ was commanded by King George as Prince George, while the _Scylla_, under Captain Percy Scott, gave birth to the “dotter,” and the “gunnery boom” which followed. In 1904 and onwards seven of them, scrapped from regular service--the _Latona_, _Thetis_, _Apollo_, _Andromache_, Iphigenia, _Intrepid_, and _Thetis_--were totally or partially disarmed and converted into mine layers. [Illustration: SECOND CLASS CRUISER OF THE NAVAL DEFENCE ACT ERA. NOW CONVERTED INTO A MINE-LAYER] The remaining eight cruisers of the Act--_Astræa_, _Bonaventure_, _Cambrian_, _Charybdis_, _Flora_, _Forte_, _Fox_, and _Hermione_--were increased in size up to 4,360 tons, and given a couple of extra 4.7, and 18-inch in place of 14-inch tubes. Instead of their 4.7’s being mounted in the well amidships, they were placed on the upper deck level, a much better position in a sea-way, but they never proved themselves quite such good ships for their size as did the earlier type. They served to illustrate the general rule that slight improvements on a design are rarely satisfactory, and that while every staple design has its defects, it is extremely difficult to remove one drawback without creating another. Moreover, such improvements invariably cause increased cost, and an essential with the small cruiser is that she shall be cheap enough to be numerically strong. Four _Astræas_ cost as much as five _Apollos_. They were rather more seaworthy, but no faster--if as fast. The total broadsides obtained were only _one_ 4.7 more and _two_ 6-inch _less_.[15] A considerably greater possible bunker capacity was obtained; but the normal supply (400 tons) was the same for both. In the British Navy, in 1908–11, a precisely similar thing obtained. It was probably inevitable. In the German Navy, between 1897 and 1907, displacement for small cruisers rose from 2,645 to 4,350 tons, with practically the same armament. But here the horse-power rose from about 8,500 or less to 20,000, and designed speeds in proportion, from a twenty-one knots (not made) to a 25.5, which, on trial, turned out to be 27,000 I.H.P. and over twenty-seven knots. Here, however, there was a definite aim--increased speed, with only trivial improvements in any other direction. With similar British cruisers the defect has invariably been “general improvements” on what the original design _might have been_ if plotted a year or two later than it actually was. There is no question--or very little--but that Germany in its ultra-conservative policy gauged the situation better than any British Admiralty ever did till just before the war.