GOETZ AND PREINING 
Fic. 6—Spiral scale pattern used on transparent foil for superposition 
on deposit; the scale of the concentric circles indicates the channel length 
L at their intersection with a spiral 
Microscopie counting procedures by such 
means have as a lower limit the statistical un- 
certainty, depending on the field area counted; 
as an upper limit the densities where the parti- 
cles start to overlap and are no longer distinguish- 
able individually. Hence the tolerance for the 
variation of Z is about 100 fold for the 1- to 0.1 
p-diameter range for A, = 10~* cm’. 
In spite of its independence from size, shape, 
and optical properties of the individual particle 
(also from accidental contaminations) the substi- 
tution of the counting method by surface pho- 
tometry is desirable for many applications, that 
is, the determination of the scattering intensity S 
from a defined small area on the channel deposit. 
This method introduces three variables absent 
in the counts: (1) the scattering power of the 
surface (foil) supporting the particles; (2) the 
variation of S with the particle size d; and (3) 
the scattering capacity specific for the particle 
type tested (refractive index, color, albedo, etc.). 
Since these variables can be controlled ade- 
quately in many cases the photometric analysis 
of foil deposits proved practical and led to the 
construction of a micro-optical foil analyzer, de- 
scribed in detail elsewhere [Preining and others, 
1959]. It represents in principle a substantially 
enlarged stage of a stationary microscope on 
which the foil is moved so that the field Ay , as 
defined by the optical system, travels accurately 
along a pre-selected spiral track within the bor- 
ders of the deposit. A specially designed ocular 
divides the light beam of the objective between a 
sensitive photocell and an eye piece, hence ren- 
ders A, simultaneously accessible for selection, 
focusing, and control counts. The close correla- 
tion of S(L) and Z(L) is thus effected as long as 
the constitution of the deposit is not too in- 
homogeneous. 
The derivation of the functions C(d) and M(d) 
of polydisperse aerosols requires the distribution 
of a strictly monodisperse aerosol along the chan- 
nel, that is, Z;(L) for 0 < L < Ly, to be known. 
The experimental production of such spectra 
meets, however, with certain difficulties. Al- 
though deposits resulting from aerosols made by 
nebulizing highly diluted suspensions of mono- 
disperse latex particles show a sharply defined La 
value (due to a defined minimum particle size d), 
such deposits are far from being monodisperse 
throughout. The presence of larger particle sizes 
is apparent from the occurrence of one or more 
distinct discontinuities of the deposit density at 
one (or more) points La’ < La. In other words, 
