OF THE UKITEU STATES. 387 



iree view of the country, must be made early in the spring 

 or late in the fall ; for the naked woods will admit many 

 views hidden in the summer by tlic branches and leaves, 

 which may be altorwards cleared away lor the work of the 

 actual triangulation. Such was for instance my case with 

 the base line itself and several important triangle points. 



This geometiical view of a country is by no means with- 

 out didicultics, and requires great attention and a kind of 

 geometric eye, as the accuracy of the work depends much 

 on tiie favourable system of triangles which is projected on 

 these data. 



The most favourable season for the actual triangulation is 

 during spring and summer, wjien the length of the mornings 

 and evenings will afford the longest time of favourable illu- 

 mination for observing, which at tiie noon elevation of the 

 sun is inadmissible. In this climate this time is often inter- 

 rupted by the frequent rains in August. 



In the fall, the atmosphere of low countries, particularly 

 the sea shore, becomes very untransparent, and the time of 

 the day favourable for accurate obseivations so limited, that 

 it can hardly ever be expected to take a full series of angles 

 consecutively. For instance, from my station at the east 

 end of the verification base near Gravcsend, in December, 

 1S17, the large lighthouse otV Sandy Hook, oidy about niiie 

 miles distant, was never visible a whole hour in the morn- 

 ing or evening: and as soon as the ground was warm in the 

 morning tlic signal at the west end of tlic verification base, 

 near five miles off, moved about irregularly in a circle of 

 about one minute. 



Tiic winter will be fully employed in tlie final calculations, 

 projecting the triangles upon paper, and oilier labours rela- 

 tive to the last summer's operations and prejiaratory to those 

 for the next summer. This will occupy more persons than 

 the actual observations have occupied in summer; as it is 

 proper that all actual calculations be made at least doultlc, 

 and by different persons. The veritiration, comparison, and 

 combination of tlie results will take up much time. 



