28 | Haakon Schetelig. [No. 8 
towards the end of the development, but observed already at an 
earlier stage. I proceed first to treat the FEastern forms which 
preserve best the original character of the type. 
3. Development of the erueiform brooches in the Eastern 
parts of the Scandinavian Peninsula. 
The Swedish archaeologists, in their description of the typology 
of the erueiform brooches, assert as å general rule that the two 
knobs which were at first placed upon the ends of the axis of the 
spring-coil, gradually lost their suggested practical destination — 
to keep the axis and the coil in 
their proper place — and became 
mere ornamental parts of the broo- 
ches. It seemed then natural to avoid 
the original and more complicated 
arrangement of these knobs by east- 
ing them in the same piece as the 
main part of the brooch,') and this 
change, which removed all practical 
difficulties respecting the construe- 
tion of the pin, was just the be- 
ginning of the pure ornamental 
development of the form. Generally 
speaking, this pronouncement very 
well corresponds with the facts 
known; it must be understood, how- 
ever, that the rule is at no time 
without exceptions, that the trans- 
formation took place much later in England than in the Scandinavian 
Peninsula, and that in Denmark traces of it are not found at all. 
It must be counted, however, among the most important points in 
1) Å similar change had taken place already at an earlier stage of 
development respecting the knob placed at the top of the brooch. In the case 
of specimens belonging to the prototype of the cruciform brooches also the third 
knob is made separately and fixed upon a projection continuing the loop of 
the axis, figs. 17, 19, and 22. The same detail is observed in å brooch of the 
half Roman form (Tromsø Museums Aarshefte 1904, p. 247, fig. 1) and is even 
preserved in some cruciform brooches with fully developed head-plate (for in- 
stance OC. 7752, Ab. 1876, p. 72, from Østerhus, Landvik pgd. Nedenes.) 
