He swam on toward the bank. Eventually it loomed up ahead of him. Hands reached down to help the lieutenant, and Blade scrambled up after him.

Blade formed the survivors into a rough defensive perimeter and settled down to wait. There was nothing they could do with his rifle and the Uzis except defend themselves, and even that might be a problem if the dragons came at them in force.

They saw no dragons nearer than the far bank of the river. Gradually the roars and bellows and screams of raging and dying dragons faded away. They began to hear the sound of boats coming up the river and helicopters flitting low over the trees, searching the area. It was one of those helicopters that found them half an hour later, and one of the boats that took them away to medical care, dry clothing, and hot tea.

Blade had cuts and bruises and a sore wrist. The doctors bandaged the wrist and let him go. He was in excellent shape to take part in the staff conferences that began meeting the next day to answer the burning question:

What Do We Do About the Dragons of the Red Flames?

The night's work had been a roaring success-two hundred dragons dead, and only light losses in Englor. But they'd been able to lay a near-perfect trap for the dragons, and that might not happen again. So last night's success really proved nothing.

It was possible to stand on the defensive. The east coast of Englor could be lined with radar stations, antiaircraft weapons, and soldiers, until few dragons could land safely or live long enough to do any damage.

It was also possible to attack the bases in the mountains of Nordsbergen, where the dragons were kept in great prefabricated domes (the ones Blade had thought might be for radar sets) until it was time to launch them on their flights across the Nord Sea. A steady bombardment from the air could kill a good many dragons and make the bases useless.

Those were the two most popular ideas. R, Blade, and Rilla were at first the only people supporting a third and much bolder proposal. They suggested flying a commando force straight into Russland aboard the VTOL assault transports. Such a force could destroy the breeding facilities; pens, and laboratories. It could kill or capture most of the key people in the whole Red Flame genetic-warfare program. At one blow it could end the threat of the dragons and set the Red Flames back ten years in their program to breed monsters for their war against Englor.

It would certainly be a bold stroke-too bold, in the opinion of too many high-ranking civilians and military men. Even R was pessimistic at first about getting his plan adopted. Then suddenly it acquired two high-ranking supporters.

One supporter was the field marshal commanding the Eighth Army in Gallia. He pointed out how many men and weapons would be needed to effectively defend Englor against the dragons. If that much strength was to be tied down on home-defense duties, he could not guarantee the survival of the Eighth Army in the face of a Red Flame attack. If a passive defense of Englor was to be adopted, he would respectfully request to be relieved of his command.

The other supporter for the commando raid was the air marshal who led Bomber Command of the Imperial Air Force. Attacking the Nordsbergen bases, he said, would commit bombers to repeated strikes against targets that would be more heavily defended each time. The losses would mount steadily. If he was called on to send his bombers on such missions, he would not take responsibility for keeping Bomber Command an effective force for operations against Russland. In such a situation he also would ask to be relieved of his command.

Both the field marshal and the air marshal were officers of long service and high distinction who had the ear of the Emperor himself. Soon the word went around that His Imperial Majesty was in favor of launching a direct attack on the breeding base in Russland.

After that, R and Blade had no shortage of supporters.

Their target was vulnerable for several reasons. First, it lay in the far south of Russland, where there were few enemy air bases and no real radar network. It was, however, less than an hour's flight from the sea. Within a hundred miles off the coast were a number of islands with plenty of room for the assault transports to land. Imperial submarines towing flexible fuel tanks could temporarily establish a secret base on one of those islands. From the islands a fully loaded assault transport could easily reach the breeding base, land its men, and bring them out again.

Second, the base itself was weakly garrisoned. The Red Flames preferred to rely on its isolation to protect it. The garrison consisted of picked Security troops, but only about four hundred of them. They were also dispersed all over a facility that covered several square miles. A heavily armed mobile force landing from the air should have no trouble crushing the garrison.

Third, the dragon pens were uniquely vulnerable. All the thousands of dragons lived in caves on either side of a deep canyon near the laboratories. They could move about freely on the bottom of the canyon, but they could reach the surface only through a few narrow tunnels.

At the head of the canyon stood a high dam. It lowered the river until the entrances to the caves in the canyon walls were above water. It also provided electric power for the whole base.

Behind the dam lay a deep lake, several miles long and hundreds of feet deep. If the dam were blown, the waters of the lake would go roaring down the canyon, submerging the lower entrances to the dragons' caves. If at the same time the upper entrances were blown in, all the dragons penned in the caves would be trapped. Long before the river went down or the tunnels were dug free, they would suffocate, to the last dragon.

The demolition would need a large quantity of high explosives, and it would have to be precisely placed by skilled men. But both were available, both could be carried to the breeding base, and the job could be done. If it was done, the dragons of the Red Flames would not trouble Englor again.

So planning for the raid began, and the arguments began soon afterward. For example, the assault transports were highly secret. It was imperative that they be fully rigged for demolition, so that if one of them crashed or couldn't take off, it would not fall into Red Flame hands. But the assault transports were also large. Enough demolition charges to thoroughly destroy one would weigh a good deal. The weight would cut into its payload for the raid. Where was the balance point between payload and precautions, if there was one?

The planners argued over large questions, small questions, and questions that Richard Blade found it hard to believe grown men could take seriously. At times it seemed that the arguments would go on until the assault transports were all obsolete and all the dragons were dead of old age.

Eventually a plan emerged. Nine assault transports would carry deep into Russland a force of six hundred men, divided into three Groups.

The Battle Group would land at the laboratories and the breeding pens. Most of the men would be riding lightweight motorcycles, so that they could move faster and carry more ammunition. They would be supported by a number of light armored vehicles with guns and rockets. Their mission was to wipe out the garrison, capture or kill the whole staff of the base, carry off everything that could be carried off, and destroy everything else.

The Demolition Group would land at the entrance to the tunnels and around the dam. The tunnel entrances would be blown in first. Then explosives would be lowered down the inner face of the dam and detonated. The pressure of the water would do the rest.

Finally, a small Blocking Group would hold the road and railroad that led out of the base to the nearest enemy garrison. They would be able to do most of the job by blowing up two bridges across small canyons.