formation which Elie De Beaumont considers as certainly belonging to the lias; and at Pouilly in Auxois, by Messrs. Bonnard and Beaumont in a bed of arkose of the same epoch of formation; both of which go to prove that this plant is one of the most important characteristics of the lias, and it is probable that it will appear again in more localities where this formation exhibits itself."*

Nine years later, Prof. Bronn, in his Lethæa Geognostica, modifies these statements of Brongniart as follows; "as yet no determinate position has been found for the sandstone of Hoer in Scania, which Adolph Brongniart includes in his common group of keuper, variegated marl, and lias, and Clathropteris meniscioides appears therein as well as in the Lias sandstone of La Marche in the Vosges, along with Equisetum columnare, which latter sandstone formation however (that of the Vosges) as it appears belongs to the keuper, in which Alberti again notices this fern in Neuen Welt hear Basle."+

In 1849, in his Index Palæontologieus, Bronn places the Clathropteris both in the Lias and the Keuper.‡

1.

W. Granite. Sandstone.

Trap.

Sandstone.

Granite. E.

a, Southampton lead mine; b, Easthampton Seminary; c, locality of Clathropteris; d, Mt. Tom; e, locality of foot-marks; f, Connecticut river; g. Mt. Holyoke Seminary; h, Hadley; i, Granby; j, Rock Rimmon.

The Clathropteris of East Hampton is found in a coarse reddish sandstone, on the west face of Mt. Tom. The upper part of this mountain is trap, beneath which the sandstone crops out with an easterly dip of about 25°. The sandstone has a southeasterly dip across the whole of the Connecticut valley. That east of the trap range is made up of finer materials, and is of a more slaty character than that on the west. Bassett's Quarry, where this fern occurs is somewhat west of the middle of the valley, as may be seen in the annexed section. In this section lately measured by my father-Dr. Hitchcock of Amherst College-it is found that the thickness of the sandstone east of Mt. Tom is more than 8,000 feet, and that on the west, or below Mt. Tom is nearly 5,000 feet. Supposing one half of this thickness to be accounted for by original deposition on an inclined surface, there will still remain a thickness of some thousands of feet both above and below the locality of the fern. A thickness equally great for the sandstone of the Connecticut valley in measuring another sec*Brongn. Veg. Foss., p. 131 and 132. Lethaæa Geognostica, vol. i, p. 140. Index Palæont., vol. ii, p. 22.

tion across the valley at Turner's Falls, 30 miles north of East Hampton.

Fig. 2 exhibits portions of several fronds or perhaps only pinnules of the East Hampton fern. Some of these pinnules-although much broken at their extremities-are nearly a foot long, and 4 inches wide, so that probably they were originally at least one foot and a half in length.

The pinnule on the left hand side of the figure, shows the delicate secondary veins, which are usually obscure and therefore this fern has not heretofore been recognised. I have presented to the Cabinet of Amherst College, a large specimen from E. Hampton containing numerous isolated fronds, and in one place showing a large number most distinctly radiating from a center, like the tree ferns of tropical regions. These radiating fronds are broken off at their extremities, being only 4 to 6 inches in length. The specimen which is figured measures nine by thirteen inches, and is in the Cabinet of the Williston Seminary at East Hampton.

In the Cabinet of Amherst College is a fine specimen of a radiating Clathropteris (that is its apex) from the quarry of Roswell Field in Gill, Mass., and although nearly one quarter of the circle is lost, yet as many as 17 distinct fronds can be counted radiating from one stem. Although the reticulated character of this specimen is rather obvious, it was not noticed till the more recent specimen from East Hampton was discovered.*

In the same cabinet is another obscure specimen of the Clathropteris from the banks of Connecticut river in Montague, two miles southwest of Field's quarry, a sketch of which is given in my father's final report on the Geology of Massachusetts, vol. ii, p. 452.

The specimens from Gill and Montague occupy nearly the same position in the sandstone as the one at E. Hampton, although they are on the east side of the trap range.

Conclusions.-The above facts make it almost certain that a species of Clathropteris occurs in the sandstone of the Connecticut valley not far from its center, measuring across the strata, and near to the interstratified beds of trap both above and below. Now since this fern is found in Europe only in the upper part of the Trias, and the lower part of the Lias, it is very probable that it occupies the same geological position here. If so we ascertain the existence of a zone of rock in the Connecticut valley not far from the junction of the Lias and the Trias. And since two measurements of sections across this valley show a thickness of sandstone strata both above and below this zone thicker than the Lias and the Trias of Europe, the probability seems very strong, that the equivalents of both of these rocks exist here, and not improbably some others both older and newer.

If we can rely with confidence upon this geological zone, it will form a convenient starting place for tracing out other older and newer formations.

It will be seen that while the above conclusions sustain the opinion lately advocated with much ability by Prof. W. B. Rogers (Am. Jour. Sci. Jan., 1855, p. 123), that the Lias sandstone exists in the valley of the Connecticut, it makes the opinion also probable-long since advanced by my father that the Trias also exists here.

The radiations in the specimen above noticed appear to be distinct and entire fronds, and so resemble those on fig. 2, as to make it probable that the latter also are fronds and not pinnules. * I have reason to suppose that this specimen has been mistaken for a Zamia.

SECOND SERIES, Vol. XX, No. 58.—July, 1855.

4

ART. III.-On the Periodical Variations of the Declination and Directive Force of the Magnetic Needle; by W. A. NORTON, Professor of Civil Engineering in Yale College,

(Continued from vol. xix, page 211.)

In the calculations made on pp. 207, 208, of the relative effects of the ecliptic and radial currents, no account was taken of the possible effects of residual currents, that is, of the gradually subsiding currents which may succeed those which are directly developed by the sun's action. I now propose to inquire into the nature and extent of their action, to make due allowance for it, and to apply additional numerical tests to the theory under examination. But I would previously remark that there is no occasion to distinguish the action of the primary from that of the secondary ecliptic currents, so called, at the equinoxes, in considering the entire effect in the interval from 6 A. м. to noon or from noon to 6 P. M. (See p. 205.) It is only when we are comparing the hourly variations that this distinction need be made. In addition to this qualification of certain statements made on p. 205, it should be observed (although the correction is of little consequence) that the secondary current, previous to 6 A. M., is in fact equally inclined to the meridian at the two equinoxes.

If we confine our attention to the entire semidiurnal interval just mentioned then the ecliptic and radial currents tend, both in the forenoon and afternoon, to deflect the needle in the same direction, at the autumnal equinox, but in opposite directions at the vernal equinox. Disregarding, for the present, the residual currents, let r denote the semidiurnal effect of the radial and e that of the ecliptic currents at the equinoxes; then taking the data from Table I, (p. 194,) we have, for the forenoon, r+e=8'65, r-e=5'67, and therefore r = 7'16 and e=1'49. For the afternoon, r+e=7'89, r-e2478; hence r5'33, and e=2'55. The two values of r differ from each other because of the action of residual currents. As a first approximation to the determination of the semidiurnal effect of the residual radial currents, let R= diff. of effects of the primary radial currents set in circulation at the beginning and end, respectively, of either the forenoon or the afternoon interval, and = effect of the residue of all the radial currents developed during either interval; then R+x=716, R-x=5'33, R=6'24, x=0'·92. Making the same calculations with the data from Table II, (p. 208,) we have, for the forenoon, r=7'59, e=1'91; for the afternoon, 2/03+1/91 r=4'50, e=203: also R 6'04, x=1'54. 2

e=

=

197. The average of the two values of e found above, from Table I, is 202. The difference of the two determinations is only 0.05.

At the summer solstice, we have, for the forenoon R+(e+x) =969, for the afternoon, R− (e+x)=3′·80; whence R= 6'74, and e+x=294. The value found for R at the equinoxes is 6'24, and of x, 0·92; hence to find r at the summer solstice, we have the proportion 6/24:674:: 0'·92 : x = 0'99. Thus e=294-x-1'95. If we make use of Table II, we obtain R 701, x=1'79, e=2'95-x=1' 16. At the winter solstice, in the forenoon, R+x-e=1'15, in the afternoon R-x +e=1'62; whence R=138, e-x=0'23. To find x, 6/24 :14:092: x=0'21. e=0/23+x=0'44. Table II gives R= 2'56, x=0·65, e=x-0'23=0'42. It is to be observed that at the two solstices the ecliptic currents cross the meridian at right angles, at the hour of noon, and that therefore the values of e just determined show the effects of the ecliptic currents on the declination at 6 A. M. or at 6 P. M. At the equinoxes these currents have but little if any sensible effect on the declination at 6 A. M. and 6 P. M., but cross the meridian obliquely at noon, and therefore the value of e at the equinoxes shows the effect of the ecliptic currents on the declination at noon.

Having obtained these first approximations let us now proceed. to a more minute investigation;-taking into account the residual ecliptic as well as radial currents.

Let R= deflection, in the semidiurnal interval, that would result from the variation in the intensity of the radial currents. E effect, during same interval, of the variation of the ecliptic

currents.

r= effect at noon of the residue of the radial currents developed during the forenoon.

e= effect at noon of the residue of the ecliptic currents developed during the forenoon.

r effect at 6 P. M. of the residue of the radial currents developed during the afternoon.

=

e' effect at 6 P. M. of the residue of the ecliptic currents developed during the afternoon.

mr portion of residual radial currents at noon that pass off in the afternoon.

ne portion of residual ecliptic currents at noon that pass off in the afternoon.

r=r′ (nearly); e=e' (nearly).

At the Equinoxes; in the forenoon, we have (from Table I); Aut. Eq., R+r+E-e=8'65; Ver. Eq., R+r-E-e=5'67. Hence E=149, R 7'16+e-r. r>e, whence R <7-16. For the afternoon the equations are, Aut. Eq., R+E+mr-r'-ne-e' = 7/89; Ver. Eq., R-E+mr-r'-ne-e'=2'78. From which we obtain E=2'55, R-5'33-(mr-r'-ne-e')=5'33+r(1-m)+ e(1+n), (nearly). No account has been taken of the residual