But I got away with it only so long. Superiors soon saw that I was a good warrior but not a hero. As they are prone to do, they got the best out of me, then consigned me to this duty of wet-nursing cadets. I will not say I am bitter about the assignment. I gain from it a certain amount of satisfaction. And it takes planning, too. I determine early on which cadets have the potential and which will undoubtedly be reassigned to the lesser castes. The warrior caste demands near-perfection, and the only way to obtain it is by winnowing down a sibko to its two, three, or four best cadets to compete in the final Trial of Position. Occasionally I have seen a sibko produce more than five—that is only logical, considering that the gene-pool contribution in some sibkos is nothing short of spectacular—but it is rare to see that many reach the last stage. It is estimated that half of the cadets fail this final trial, too. I am proud that none of the sibkos I have supervised has ever failed to produce at least one successful warrior.

There are those who criticize the warrior program as it is practiced here on Ironhold, those who have an economical turn to their minds. They say that we produce too few warriors, that our armies grow too slowly, and BattleMechs will be left gathering dust in underground caches. Their arguments become poignant when they discuss actual wars, where warriors may be killed at a faster rate than we are turning them out here. Yet we do produce many more warriors than anyone might expect, what with all the training units spread out over Ironhold. I am, after all, only one of more than a hundred falconer commanders in charge of training units, and each of our units processes at least twenty sibkos. I am overseeing twenty-six sibkos at present, ranging from newly arrived contingents to the sibkos whose members have been whittled down to their survivors. I believe we are, in fact, shipping out warriors at an astonishing rate, given the demanding, harrowing, and long training program they have to get through. Nicholas Kerensky would, I think, have been satisfied, and undoubtedly proud, of our accomplishments. The warriors we produce do tend to validate his genetic programs, starting with the gene pool from which most of us who qualify have emerged. His theories of eliminating the worst traits, those that interfered with the skills and thus the success of a warrior, and transmitting the best traits, those of our most skilled and wondrous warriors, to sibling companies have been proven over and over. And here on Ironhold, we carry on the theories by taking these products of the genetic program and doing our own eliminations. The end result is that we train the best warriors humanly possible. That is the way of the Clan and the wonder of the genetic program.

I am not prone to nostalgia (a negative trait if there ever was one), but I sometimes think of my own sibko and our days in training. We were a rough group, unlike some of the sibkos in training now, and at least half of us disposed of the other half before we got down to the serious business of molding ourselves into cadets worthy of testing. My first time out in a stripped-down 'Mech, I killed one of my sibkin. Looking down at his corpse, I wondered if we had ever been close. I walked away from the body with no regrets. And regret has not been a part of my arsenal since.

The Mattlov/Pryde sibko has reached its final stages. Five youngsters remain, including Cadet Aidan, whom I have observed carefully ever since noting such a strong Mattlov resemblance. They are nervous, eager, almost ready to attain that curious psychological blend of individual and machine that occurs when a warrior is at one with his 'Mech. I have tried to explain this feeling to many people over the years, but few non-warriors have even approached understanding. Even some warriors have claimed ignorance of this phenomenon, saying it is just the feedback from the neurohelmet that creates the illusion of oneness. But it could not be merely the neurohelmet. I fought countless battles with my headgear damaged and lost not an iota of my connection to the machine. I cannot imagine any warrior being successful in a 'Mech without having the sense of it as a living being for whom he or she is the driving force, the brain. But even that does not express it. The meld feels almost like a joining of metal and circuitry with the skin and innards of the pilot. Drivers of vehicles have told me they have often felt the same way about their machines.

Looking at the roster of Aidan's sibko, as I am doing now, I am impressed with the achievements of this quintet of survivors. Still, I suspect that one or two of them might not make it to the last test. Cadet Peri, whose intellect nearly matches that of the top student, Cadet Marthe, still lags appreciably in her mechanical skills. It may be dangerous to allow her to go much further. She will be useful in another caste. No sense wasting a life for purposes of bravado. I must speak with Falconer Joanna about Peri.

Of the rest, Marthe and Aidan have the superior skills, although only Marthe really knows it. I can see the doubts in Aidan's eyes, the residual effects of the riding that Joanna gives him. She has done everything to take away his confidence, to break him. But he keeps coming back, rising to his impressive full height as he did when she fought him so hard on that first day long ago when the sibko first arrived on Ironhold. Resilience seems to be his special talent. Yet, there is a weariness in him that worries me. I have told Joanna not to be so harsh with him, but she is adamant. She does not believe in the theory that says a training officer should go from being disciplinarian to mentor. Indeed, she insists that being kind to a cadet, even one who seems destined for success, gives him or her a certain slack that can affect everything from concentration to timing. Perhaps she is right. With my irritability and sudden temper, I would make a poor mentor. Yet, other officers who advocate the milder approach also turn out successful warriors with their methods.

It often seems strange that our training goes against the grain of military training as practiced in past eras. In earlier times, the idea was to take separate individuals and mold them into a unit that would work and fight efficiently together. As I understand it from some readings, the process consisted of conditioning the minds of the trainees. Any trace of individuality was removed in favor of group thinking so that the military unit would be united. In our approach, we travel an opposite route. We take a group that is united, a sibling company or sibko, and break down its unity. We even set them against each other, as we did with this particular sibko. And why? So that we can turn them into individuals, give them the singularity necessary to the character of a BattleMech pilot. Oh, we realize the necessity of unity in battle, but that comes later. Assigned to a Star, the individual relearns the unity of the sibko, this time with new companions. And, some say, it is a new and better unity, one that adapts to new warriors coming in to replace dead or departed ones, and to new units. The old team tussels of the sibko seem primitive when compared to the feisty concord and loose harmony of a genuine fighting unit.

It is hard to predict what will happen to the other two cadets in the Mattlov/Pryde sibko. The short one, Bret, is a battler, all right, and reasonably intelligent, but he is more bravado than skill. He might make it. He is certainly out to prove himself and will accept nothing less than victory, a quality we always say is necessary for a warrior.

The other one, Rena, does not quite seem like warrior material. She was once overweight and still moves with some of her former clumsiness and heaviness. Yet she has courage and a tenacity toward surmounting obstacles, so she may surprise us all.