to 
he 
to 
ry 
aS 
positioning was known to within 0.01 mm; because of this they 
could also be moved in relation to the vertical by means of 
special ly-designed micro-metric screws. Moreover, the wires 
themselves were held tanght by a 5 Kg weight immersed in a 
container of oil to absorb the vibrations. 
In Fig, 1 there is a plan and section diagram of the equi 
pment just referred to, and the relative positioning of the 
individual pegs. 
To determine the spatial co-ordinates of these, ‘we have 
had to work solely with a series of distance measurements bet 
ween the brass pegs using a method already described in the 
afore-mentioned publication (3). In this way we set up a spa- 
tial frame independent of the model to be studied (which obviou 
sly had to be placed in a vertical position). Previously, tests 
had been done outside our laboratory using an independent refe 
rence system, implemented with "Johnson" blocks mounted on a 
reference table; in this case the model studied was placed di 
rectey on the table. 
Given the impossibility, at the moment, of setting up this 
equipment in our laboratory, we have not been able to experi- 
ment with it in depth, though it seems to us to be the more ap 
propriate and more accurate solution. 
The objects examined were: 
a) a model of a part of a car body, to a scale of 1.1, made of 
hardened resin; 
b) a model in sheet metal, also to a scale of 1.1, of the left 
front mudguard of a Fiat 125. 
The dimensions of model (b) are: 110 cm x 60 cm, with a 
maximum width of 13 cm. The sheet metal model was tightly fixed, 
with a number of specially-placed small connecting-rods, to a 
plate made up of a panel of two joined sheets of wood, hardened 
and pressed to a thickness of exactly 2.5 cm and having dimen- 
sions of 130 cm x 80 cm.