Since the author’s arrival in New York for the purpose of publishing his outlines, the third and fourth volume of the Cosmos has been placed in his hands, containing the latest uranological discoveries and speculations. It is now more than twenty years since he began to investigate the subject he has treated of, and fifteen since he first announced to the world, that he had satisfactory evidence of his theory being true. Luckily, perhaps, he has been cut off from the great streams of knowledge; and he may confess that it was with pardonable feelings of gratification that he discovered in 1853, by the acquisition of the two first volumes of the Cosmos, that the philosophic mind of Humboldt had also pondered deeply on the planetary peculiarities of size, density, distance, inclination of axes and eccentricities of orbits, without eliciting any satisfactory relations.

From the tenor of the third and fourth volume of this learned summary of scientific knowledge, it is evident that the question of a medium filling space is more and more occupying the learned world; but the author is unable to discover any consistent theory respecting it. The increasing interest attaching to it, however, is evidently preparing the world for some radical change in preconceived views. The explanation given by this present theory to many prominent phenomena, is so totally contrary to that of the learned world, as to leave it untouched by anything yet advanced. What the fifth volume of the Cosmos will contain, is not yet known in this country, neither has the author been favored with any glimpse of the progress of science as developed before the British Association; he supposes, however, that he yet stands alone in the position he has defined.

As a question of practical importance, the reader will find in the work cited, the various opinions of the temperature of space. Both Fourier and Poisson regard this as the result of radiated heat from the sun and all the stars, minus the quantity lost by absorption in traversing the regions of space filled with ether.[51] But why should we regard the stars as the source of all motions? Why cannot physicists admit the idea of an infinite space filled (if we may use the expression) with an infinite medium, possessing an unchangeable mean temperature long before the formation of a single star. A star equal to our sun at the distance of Sirius, would give about one million of million times less heat than our present sun, which is only able to give an average temperature to the whole globe—about twenty degrees above freezing—then let us remember that there are only about fifty stars of the first and second magnitude, which give more light (and by analogy heat also) than all the rest of the stars visible. Such labored theories as this of Poisson’s is a lamentable instance of the aberrations of human wisdom.

We would also call the reader’s attention to a late conclusion of Professor Dove, viz.: That differences of temperature in different longitudes frequently exist on the same parallel of latitude, or, in other words, are laterally disposed. This may be thought adverse to the theory, but it should be borne in mind that the annual mean temperature of the whole parallel of latitude should be taken when comparing the temperatures of different years.

Another fact cited in the Cosmos apparently adverse to the theory, is the idea entertained by Sir John Herschel, that the full-moon dissipates the clouds. This question has been fully examined by Professor Loomis before the American Association, and he concludes that there is not the slightest foundation for the assertion—taking as data the Greenwich observations themselves.