430 On Acriat Navigation. 
my former paper. I do not mention this circumstance with the 
view of questioning the originality of Sir George’s idea, but 
merely because it may be usefal, in the prosecution of further 
experiments, to know what constructions have been already ex- 
amined, and what difficulties have impeded their execution. 
It appears a very obvious improvement to make the balloon 
itself furnish the required oblique surface as well as the moving 
power. In proceeding to submit this to experiment, the first ques- 
tion that offered itself was, what should be the shape of the bal- 
Iqon in order to obtain the. oblique surface most conveniently? 
From previous experiments, I was well acquainted with the dif- 
ficulty of distending the balloon by framework however lightly 
made. I therefore so ught some shape which should not require 
such assistance, but be such that the balloon coukl not assume 
any other figure without diminishing its capacity. The shape 
of the first balloon. censtructed with this view, was that of two 
flattened hexagonal pyramids united by their ‘bases: The bal- 
loon when inflated did not perfectly retain the above figure, but 
became rounded at all the ¢orners and bounding lines. The 
aperture for the fire was not at the apex of the inferior pyramid, 
but on one side of it, and a weight was attached which might be 
adjusted so as to keep the balloon at the requisite obliquity. 
The balloon was elevated in the room where I made the experi- 
ments detailed in my former paper, but no sensible effect was 
observed. Being afterwards launched in the open_air with a 
small side sail attached, it assumed the same rotatory motion as 
a balloon furnished with an inclined plane. 
In this experiment much inconvenience arose from the constant 
tendency of the hot air to ascend to the higher extremity of the 
balloon, which was consequently always endeavouring to attain 
a perpendicular sositionn It is much to be feared that this cir- 
eumstance would be a material obstacle to the success of bal- 
loons constructed after this plan on a large scale; which con- 
sideration inclined me to prefer the original method of suspend- 
ing the plane at some distance below a spherical balloon. The 
experiments on the comparative advantages of the two construc- 
tions are however at present too few and imperfect to warrant a 
decision as to the superiority of either of them. 
Sir George in his last paper gives a description of a hydrogen 
gas balloon furnished with wings or oars, which are to be moved 
alternately up and down byasteam-engine. It appears to me that 
by a force being communicated to these wings in one direction, 
which force is alinost immediately destroyed in order to give an 
impulse in the contrary direction, a considerable loss of power 
must_be occasioned. I would therefore beg leave to suggest 
that a large wheel with oblique vanes (like the fly of a wae 
jac i) 
