instant of taking the rod reading, which fact is far more objectionable than a slight parallax. Also, in reversing the level tube the prisms over the level must be reversed and refocused, which consumes time and is apt to disturb the instrument.

As will be seen from Table No. 1, this instrument is not nearly as delicate, in point of magnifying power and level tube curvature, as are those previously described.

d. The United States Coast and Geodetic Survey Level. This instrument, shown in Fig. 1, Plate III., was designed and made in the shops of the Survey, and represents the type of level used by this bureau of the Government prior to 1899. Some changes in this instrument, as also in the field methods used by the Survey, have been made recently, but these have not yet been made public, and hence only a passing reference will suffice to point out a few of the most striking features which are different from other high-grade levels.

The general design does not differ materially from the Swiss instrument. There is no means, either by mirror or otherwise, of observing the bubble simultaneously with the rod, which is a very serious omission. Also, there is no coarse level for approximate setting of the instrument, and hence this must be done with the striding level in two positions, which is a slow process. A novelty (supposedly) is introduced by a system of bearings or wyes, inside the principal wyes, for the purpose of raising the telescope off its supports while it is being revolved on its horizontal axis, to prevent the wear usually attending this manipulation. However, as has been intimated, and as will be brought out more clearly in another place, this is not necessary, because the wear of the collars does not influence the relation existing between the level tube and the telescope, except when the tube contact and the wye points of support are on the same circles of the telescope collars, which is seldom the case and should be avoided in all instrument designs.

e. The Massachusetts State Survey Level. This instrument, shown in Fig. 2, Plate III, was constructed by Messrs. Buff and Berger, of Boston, Mass., and used also by the United States Geological Survey. It is not really a precise level, but merely a good common level.

This instrument, in principle very much like that shown on Plate II., embodies most of the features of a precise level, but, without a striding level tube, it cannot be considered as such. The focusing is accomplished by motion of the objective, which is somewhat inconvenient to the observer. This instrument could readily be altered to become a very good level for precise or common work merely by removing the fixed level and replacing it with one that can be used as a combined striding and fixed level, as is done on the Buff and Berger plane table alidade. However, the mirror should not be omitted, and the form used in Plate II. would be preferable to that here used. The properties of the five instruments just described are given in Table No. 1.

TABLE No. 1.-CONSTANTS OF VARIOUS PRECISE LEVELLING INSTRUMENTS.

Reversible Revers, mirror 5.0 changed 5.3 7.6 12.9
striding.
over level.

1.7 to 3.4 5.0 5.8 10.8 watchglass
level.

a. Buff & Berger Level
No. 2768. Pattern of
Prof. Mendenhall

FIG. 1, PLATE III.-U. S. COAST AND GEODETIC SURVEY LEVEL.