^{page 196 note 1} It should be understood that the examples of both types 1 and 2 illustrated on pl. xxxvi are imperfect, lacking the lateral framework which supported the spring, whether single-coiled as on that from Hungary or double-coiled as on that from Szabadbattjan (fig. 2, 7). A side-piece which was believed by Hampel to have belonged to a brooch very similar in detail and dimensions to that illustrated on pl. xxxvi, a, is figured by Hampel (op. cit., ii, 31,figs.A-B). In every case these brooches when complete had a knob at the top of the head-plate and a pair or double pair of knobs at the sides.

^{page 200 note 1} Brooches are the most distinctive class of relics by which the material stability of population in a given area during the migration period can be judged.

^{page 200 note 1} ‘Ein Beitrag zur Chronologie der Völkerwanderungszeit’(Mannns-Bibliothek, 22, pp. 108 ff.).

^{page 201 note 1} Op. cit., pp. 262–7.

^{page 201 note 2} pp. 12–13.

^{page 201 note 3} §§113ff.

^{page 201 note 4} Cf. settlement of the Ostrogoths under Attila and VOL. xciv Ernac's refuge by permission of Marcian at the confluence of the Danube and Tisza. ( Thompson, E. A., A History of Attila and the Huns, pp. 153–4.)Google Scholar

^{page 202 note 1} Ammianus Marcellinus, xxxi, II: ‘… auri cupidine immensa flagrantes’; Hieronymus, Epp. lxxvii, 8: ‘ob nimiam auri cupiditatem.’ Professor Thompson in his book, which did not appear until after this paper was written, contributes much valuable information in corroboration of this trait of the Huns. They, a nomadic people,unversed in almost all manufactures, required gold to purchase not only food but also gear of every kind, even weapons (see pp. 4, 57, 161, 170, and 193–4). Incidentally he quotes Sir Ellis Minns's opinion that ‘he does not think the nomads worked metals. Metal work, if not all work, was for slaves, tributaries, and neighbours to supply.’

^{page 203 note 1} Schmidt, L., op. cit., p. 7Google Scholar; Dio Cassius, 71, 12.

^{page 204 note 1} See also Schmidt, L., Geschichte der deutschen Stämme, i, 18Google Scholar.

^{page 204 note 2} The Gothic History of Jornandes (Princeton, 1915), p. 83.Google Scholar Unfortunately the late Professor Alexander Souter did not cite this occurrence of the word in his Late Latin Glossary. His view on its meaning here would have been interesting.

^{page 204 note 3} Patsch, Carl, Beiträge zur Völkerkunde von Südost-europa, iii, 137 fGoogle Scholar.

^{page 204 note 4} Patsch, , op. cit., p. 38,Google Scholar evidently held the same view, since he notes the repatriated Sarmatae were, in 35⁸, settled as foederati, but in numbers insufficient to cover a territory that had previously been thickly populated. He adds that he considers it highly probable that at that period the friendlily disposed Victovali from northern Hungary were transplanted thither as the first German (i.e. as contrasted with Sarmatian) peoples in the Banat.

^{page 205 note 1} In the passage mentioned the Vandals in question are described as living round Lake Maiotis, adjoining the Crimea. Little warrant seems to exist for any Vandal settlements in that quarter, though it could be true of Alans, who, in 332, as Jordanes states, along with Vandals ( Schmidt, L., op. cit., p. 12)Google Scholar obtained imperial permission to settle in Pannonia. By 400 it may well be that Pannonia, too, was becoming severely overcrowded owing to other' retreats from the advancing Huns.

^{page 205 note 2} A word of caution is necessary in regard to the term Blechfibel by which type 1 is generally known to continental archaeologists. Primarily it has a technical signi- ficance, but later it is loosely used to denote a class in which the general form is preserved without distinction as to whether the ornamental apron covering the mechanicalpart of its structure is hammered from sheetmetal or is cast, an important difference, where the name Blechfibel can become a misnomer and misleading.

One consideration must also be borne in mind. The form of type I is technically simple, and, so long as it was produced by hammering with a keel running down the foot was incapable of any great development except in point of size. As Beninger has well put it, its form is starr und unbeiveglich—stiff and immobile. Appliques, as at Nagy Varad, were almost the only possible addition, and even in these there is an extraordinary conservatism, best illustrated by the acanthus frill at each end of the bow. Engraving or punching seems hardly, or never, to have been practised. The type cannot have had a very long life; 150 years seems a fair estimate. Zeiss, however, talks of 250 years, a figure which seriously over-estimates its possibilities, if one starts from the early evolutionary stages as seen at Marosszentanna.

^{page 206 note 1} Prähistorische Zeitschrift, 1912, p.157Google Scholar.

^{page 206 note 2} How great a tribe the Vandals must have been is shown by their constant participation in raids across the Danube in pre-Hun days, by their association with Radagaisus's expedition into Italy in 403, and by their loss, according to Jordanes, of 20,000 men (a figure probably greatly exaggerated) in their clash with the Franks in 406. In spite of that they were able to raid Gaul and after 409to penetrate the Iberian peninsula almost at once as deep as into Andalucia. And finally, even after quarrels with their allies and conflicts with Visigoths, to cross over into north Africa in 425, and from there to harry the Roman Empire for twenty-five years. On the question of numbers in general recorded by the various chroniclers L. Schmidt has some valuable comments in his Geschichte der deutschen Stdmme,i, 45-6.

^{page 208 note 1} Op. cit. 31-2, pl. 7, 3.

^{page 208 note 2} Album Caranda, pl. 95, nouv. serie, fig. 1.

^{page 208 note 3} Zeiss, , op. cit., pl. n, 1 and 2Google Scholar.

^{page 208 note 4} Vandal is used throughout this paper for brevity, but all the partners are held in mind.

^{page 208 note 5} That this type of jewellery, worn by the women-folk of the invading host, left little impression in Gaul is not surprising when one considers the conditions underwhich it was introduced and the shortness of the wearers' stay there.

^{page 208 note 6} Kubitschek, W., Jahrbuch für Altertumskunde, v, 32, pl. 11, fig. 5.Google Scholar It is to be noted that an advanced date for this brooch is demanded by the absence of the median ridge on the foot which characterizes the true Blechfibel. Its date is certainly later than c. 400, where Åerg places the grave, the jewellery from which covers a long period.

^{page 208 note 7} It is to be seen on the huge example of type 3 from Flamicourt, Somme ( Bull, , archeól., 1895, p. 94, pl. XXII, 5)Google Scholar here fig. 4, 7.

^{page 210 note 1} e.g. Rouise, Syrmium ( Åberg, op. cit., fig. 78),Google Scholar Gáva, Hungary, ibid., fig. 79, Úherce, Bohemia, ibid., fig. 18.

^{page 210 note 2} Ibid., fig. 101.

^{page 210 note 3} Migne, , Patr. Lat. 22, 600Google Scholar.

^{page 211 note 1} The stress which should be laid upon the term settled cannot here be too strongly emphasized. It may seem difficult, almost paradoxical, to use the word settled at all in reference to the areas north of the Danube in the fourth century. Regarded, however, in their proper aspect, many of the quarrels and disturbances among the barbarian tribes have little more significance relatively speaking than the outbursts of temper or hunger that, in the Middle Ages, gave rise to mutual attacks north and south of the Scottish border. In neither case had they the devastating, deracinating effect of the Hun invasion. Amid them the tribes must have enjoyed some settled life; Roman subsidies or douceurs must have kept many a tribal exchequer well-filled. There cannot be the least doubt that a reasonably settled life was possible and did exist. The women's graves, from which the large Blechfibeln come, are the abundant proof of it.

^{page 211 note 2} Jenny, A. von, Germanischer Schmuck, 40, pl. 15, 1–2Google Scholar.

^{page 212 note 1} Den Nordiska Folkvandringstidens Kronologi.