APPENDIX. 
423 
fire with her straw hat, threw herself on the ground near by to 
rest and smoke her puro. When the pot was near to boiling, she 
reached out her bare leg and tested the temperature of the contents 
with her toe, as a Northern cook might have used his finger. Frank 
was scandalized; but, after all, it was merely a matter of taste. 
PHOTOGRAPHS USED IN ILLUSTRATION. 
In stating that the scenes illustrated in this book are all from 
photographs, it may be added that the clearness of the atmo¬ 
sphere enables a distant view to be taken with great distinctness 
(unfortunately lost in the mechanical reproductions) even in 
minute details. The lens used for views not requiring extreme 
rapidity was the Dallmeyer single landscape,— a lens unsur¬ 
passed for its purpose ; while for architectural subjects, or those 
in motion, a Ross rapid rectilinear was generally used. The 
plates were those prepared by Allen & Rowell, of Boston, 
as usual, of the finest quality. For apparatus, the camera 
was a 5 X 8 size of the American Optical Company’s make, 
fitted with a changing box containing eighteen plates, and also 
with an attachment, arranged by the author, for making two or 
three smaller pictures on the 5x8 plate. I carried no tent, 
but changed my plates at night under a blanket, depending on 
touch rather than sight. For the stereoscopic pictures, I used 
a pair of Euryscope No. 0 lenses. The plates w r ere developed 
months afterwards, with a very small percentage of failures. In 
later journeys in Guatemala I have used plates of the 8 X 
size; but for all purposes of illustration the 4x5 size is to be 
preferred. For packing the plates I have used a strong barrel 
and cork-dust with complete success. It is a matter of deep 
regret that the method of mechanical reproduction utterly de¬ 
stroys all the beauty of the original photographs. In cases 
where phototypes are presented from ink-drawings, these have 
generally been drawn directly from a transparency which I ha\e 
made from the original negative and projected in the lantern. 
The drawings are of large size, and reduced to one quarter, or 
even less, in the phototype. This method insures at least 
accuracy of outline. 
