Ten days after their first appearance off Tsushima, the Toi effected a landing in Chikuzen and marched towards Hakata, plundering, burning, massacring old folks and children, making prisoners of adults, and slaughtering cattle and horses for food. It happened, fortunately, that Takaiye, younger brother of Fujiwara Korechika, was in command at the Dazai-fu, whither he had repaired partly out of pique, partly to undergo treatment for eye disease at the hands of a Chinese doctor. He met the crisis with the utmost coolness, and made such skilful dispositions for defence that, after three days' fighting, in which the Japanese lost heavily, Hakata remained uncaptured.

High winds and rough seas now held the invaders at bay, and in that interval the coast defences were repaired and garrisoned, and a fleet of thirty-eight boats having been assembled, the Japanese assumed the offensive, ultimately driving the Toi to put to sea. A final attempt was made to effect a landing at Matsuura in the neighbouring province of Hizen, but, after fierce fighting, the invaders had to withdraw altogether. The whole affair had lasted sixteen days, and the Japanese losses were 382 killed and 1280 taken prisoners. Two hundred and eighty of the latter—60 men and 220 women—were subsequently returned. They were brought over from Koma six months later by a Koma envoy, Chong Cha-ryang, to whom the Court presented three hundred pieces of gold.

Kyoto's attitude towards this incident was most instructive. When the first tidings of the invasion reached the capital, the protection of heaven was at once invoked by services at Ise and ten other shrines. But when, on receipt of news that the danger had been averted, the question of rewarding the victors came up for discussion, a majority of the leading statesmen contended that, as the affair had been settled before the arrival of an Imperial mandate at the Dazai-fu, no official cognizance could be taken of it. This view was ultimately overruled since the peril had been national, but the rewards subsequently given were insignificant, and the event clearly illustrates the policy of the Central Government—a policy already noted in connexion with the revolt of Masakado—namely, that any emergency dealt with prior to the receipt of an Imperial rescript must be regarded as private, whatever its nature, and therefore beyond the purview of the law.

A more effective method of decentralization could not have been devised. It was inevitable that, under such a system, the provincial magnates should settle matters to their own liking without reference to Kyoto, and that, the better to enforce their will, they should equip themselves with armed retinues. In truth, it is not too much to say that, from the tenth century, Japan outside the capital became an arena of excursions and alarms, the preservation of peace being wholly dependent on the ambitions of local magnates.

A history of all these happenings would be intolerably long and tedious. Therefore only those that have a national bearing will be here set down. Prominent among such is the struggle between the Taira and the Minamoto in the Kwanto. The origin of these two families has already been recounted. Some historians have sought to differentiate the metropolitan section of the Minamoto from the provincial section—that is to say, the men of luxury and literature who frequented the capital, from the men of sword and bow who ruled in the provinces. Such differentiation is of little practical value. Similar lines of demarcation might be drawn in the case of the Taira and Fujiwara themselves. If there were great captains in each of these famous families, there were also great courtiers. To the former category belonged Taira Tadatsune. For generations his family had ruled in the province of Shimosa and had commanded the allegiance of all the bushi of the region. Tadatsune held at one time the post of vice-governor of the neighbouring province of Kazusa, where he acquired large manors (shoen). In the year 1028, he seized the chief town of the latter province, and pushing on into Awa, killed the governor and obtained complete control of the province.* The Court, on receiving news of these events, ordered Minamoto Yorinobu, governor of Kai, and several other provincial governors to attack the Taira chief.

*Murdoch, in his History of Japan, says that in three years Tadatsune's aggressions "reduced the Kwanto to a tangled wilderness. Thus, in the province of Shimosa, in 1027, there had been as much as 58,000 acres under cultivation; but in 1031 this had shrunk to forty-five acres."

Yorinobu did not wait for his associates. Setting out with his son, Yoriyoshi, in 1031, he moved at once against Tadatsune's castle, which stood on the seashore of Shimosa, protected by moats and palisades, and supposed to be unapproachable from the sea except by boats, of which Tadatsune had taken care that there should not be any supply available. But the Minamoto general learned that the shore sloped very slowly on the castle front, and marching his men boldly through the water, he delivered a crushing attack.

For this exploit, which won loud plaudits, he was appointed commandant of the local government office, a post held by his grandfather, Tsunemoto, whom we have seen as vice-governor of Musashi in the days of Masakado; by his father, Mitsunaka, one of the pillars of the Minamoto family, and by his elder brother, Yorimitsu, who commanded the cavalry of the guards in Kyoto. The same post was subsequently bestowed on Yorinobu's son, Yoriyoshi, and on the latter's son, Yoshiiye, known by posterity as "Hachiman Taro," Japan's most renowned archer, to whom the pre-eminence of the Minamoto family was mainly due. Tadatsune had another son, Tsunemasa, who was appointed vice-governor of Shimosa and who is generally spoken of as Chiba-no-suke. The chief importance of these events is that they laid the foundation of the Minamoto family's supremacy in the Kwanto, and thus permanently influenced the course of Japanese history.

THE CAMPAIGN OF ZEN-KUNEN

It is advisable at this stage to make closer acquaintance with the Japanese bushi (soldier), who has been cursorily alluded to more than once in these pages, and who, from the tenth century, acts a prominent role on the Japanese stage. History is silent as to the exact date when the term "bushi" came into use, but from a very early era its Japanese equivalent, "monono-fu," was applied to the guards of the sovereign's palace, and when great provincial magnates began, about the tenth century, to support a number of armed retainers, these gradually came to be distinguished as bushi. In modern times the ethics of the bushi have been analysed under the name "bushido" (the way of the warrior), but of course no such term or any such complete code existed in ancient days. The conduct most appropriate to a bushi was never embodied in a written code. It derived its sanctions from the practice of recognized models, and only by observing those models can we reach a clear conception of the thing itself.