284* Notices respecting New Books. > 



ffjtst. 



only the £$ec£ of their sketches but the facility b£ the time saved in 

 taking them, will soon overpay the time employed in acquiring the 

 theory and the practice ; — that the hours spent in. the study will soon 

 be compensated by the hours saved in the field ; by means of the 

 possession of principles, and the knowing what to aim at. In the 

 conclusion of the former part, the author expressed himself as un- 

 willingly breaking off, while some parts of his subject remained but 

 partially treated of; leaving the intention of continuing the work, 

 dependent on the reception his volume should meet with. 



In the Introduction to the present Supplement, he says, " Without 

 waiting for the result, he has considered such parts to be essential 

 to the theory, and the work incomplete in its plan without the in- 

 troduction of them." 



The Supplement begins with the Field of Vision, of which the 

 author gives a more complete theory than in the first volume, stating 

 more particularly the ground of its limits; and that those limits do 

 not subsist where other writers have placed them; — showing also 

 why pictures, when reduced for the purpose of engraving, or other- 

 wise, will not produce the same perspective effect in the larger ori- 

 ginals, although the same proportions may be exactly maintained ; and 

 showing also why, in small diagrams, the Perspective, though true, 

 will strike the eye as distorted. In the chapter on the Horizon-line, 

 he examines and in some measure opposes the theories of for- 

 mer writers, and maintains that the height of the horizon in the 

 picture " has no more to do with the height or width of the picture, 

 than it has with the thickness of the wood-frame in which the pic- 

 ture is placed." He then illustrates his theory in a very clear man- 

 ner, by a series of wood-cuts, representing sea and ships with the 

 same outline and on the same scale, but as viewed from different 

 heights, and therefore with different positions of the horizon-line; 

 adopting a new term, and making a distinction between a high hori- 

 zon, and an enlarged sub-horizon or low base ; and vice versa : and 

 this hitherto irregularly discussed theory, he appears to us to have 

 placed on a footing from which it is not likely to be removed. Mr. 

 Davenport afterwards contends, against the authority of almost all 

 his predecessors on this subject, that the prime vertical line must 

 be placed equally distant between the two sides of the picture; and 

 gives his reasons why the " prime vertical must cut the horizon-line 

 in its centre, though it is not necessary that it should itself be inter- 

 sected by that line in its own centre." He endeavours to give a phi- 

 losophical rationale of the causes of these and other apparent ano- 

 malies, and of the difficulties and deceptions that occur in pictorial 

 representation; such as the difficulty of dovon-hill views ; — the de- 

 ceptive magnitudes of objects on very high edifices, as compared 

 with those at equal horizontal distances ; and, contrary to the most 

 generally received estimation, criticizes what he calls the deceptive 

 flfo-proportions of St. Peter's at Rome. He also gives some further 

 problems principally for the finding of distant vanishing points, 

 and notices the effect of the refraction of the atmosphere, of reflection 

 from the surface of water, &c. ; but we think that some examples 



both 



