When the fleet was actually in touch with a hostile fleet, in pursuit, or when expecting battle, the frigates were placed between their own force and the enemy; nearer, however, to the latter, as the essential point was to keep knowledge of his whereabouts and probable intentions. Such a position was at times extremely exposed. The frigates had to avoid equally capture and being driven and shaken off; they must keep close, yet not be caught. When engagement ensued they passed through to the off side of their own fleet, where they were dispersed at intervals abreast the main line, like the file closers of a military line ashore. Here they fulfilled one special purpose, besides others. As the fleet fought with broadsides only, its ships were ranged one ahead of the other. Consequently signals made on the masts of the admiral could not be seen always by those ahead or astern of him; but the frigates in the other line made the same signals, “repeated,” as it was said, where they could be read more certainly. But frigates did also more hazardous work. They went to crippled ships of the line and towed them into other positions, into or out of fire, and at times the admiral summoned a frigate alongside to carry a message to some part of the battle. “I noticed,” says Marryatt, in one of his novels, “the look of pride on the faces of our officers when it appeared that the loss on board our frigate was greater than that of some of the ships in the line.” For such offices it is evident there were wanted a strength and a weight which the corvette did not have. A corvette would make poor work of towing a heavy ship, and could not carry as surely the sail needed to maintain a position. At the same time it should be observed that excess of size above the requirements stated should be exceptional. In the opinion of the writer the forty-four-gun frigate in her day possessed a fighting force and a weight of body in excess of that required by the ordinary functions of her nominal class. For exceptional reasons, a few of the type were permissible in a large navy. On the other hand, it may be inferred from the long experience of the British navy, and the resultant practice, that ships of twenty-eight, twenty-four, and twenty guns, though often styled frigates, were not found satisfactory as such. In the distribution of tonnage between size and numbers, a mean must be found; and it must be added that a just mean is a very different thing from a compromise. These considerations also apply to present-day problems. EARLY SHIPS OF THE LINE