136 Proceedings of Royal Society of Edinburgh. [sess. 
experiments. Though that apparatus is satisfactory enough for 
preliminary work, and gives fairly good results, yet it is evidently 
not suited for regular everyday use; and, besides that, there are 
certain defects in it which can he avoided in apparatus specially 
constructed. If a regular examination of the dust in the air was to 
he made from day to day, it seemed advisable that I should devote 
some time to devising a more practical form of apparatus, one 
which could he more easily and quickly worked, and which could 
he managed hy any one not having a special knowledge of the 
subject. 
With this object in view I have, during this summer, devoted 
a considerable time to the construction and testing of the new 
apparatus ; and the object of this paper is to describe the improve- 
ments I have been able to make, and to give detailed working 
drawings of the different parts of the apparatus, so that any instru- 
ment-maker may be able to reproduce it. I may as well give notice 
here that the general reader is warned off, as what follows is a 
dreary desert of mechanical details, which however necessary for 
those who are going to assist in developing the investigation, yet 
contains nothing new of scientific interest. Anything new in that 
direction will be found near the end of the paper. 
The alterations and improvements will be best understood by the 
drawings given with this paper. Plate I. shows the general arrange- 
ment of the different parts as fitted up for work ; Plate II. gives 
detailed drawings of the test-receiver ; while Plate III. shows the 
apparatus for measuring the air to be tested. Most of the draw- 
ings on Plates II. and III. are to a scale of half size. 
Returning to Plate I. showing the general arrangement, there 
are five distinct parts in the apparatus — (1) the test-receiver R ; 
(2) the air-pump P ; (3) the measuring apparatus M ; (4) the illum- 
inating arrangements L ; and (5) the gasometer G. The air to be 
tested is drawn through the pipe A by means of the gasometer and 
its connecting pipes. The air on its way passes through the 
measuring apparatus M, where a measured quantity of it is taken 
and passed into the receiver R, where it is mixed with a certain 
quantity of dustless air, and saturated with water. The air in R is 
then expanded by the pump P, a shower of rain produced, and the 
number of drops which fall on a measured area are counted. Such 
