388 
ASTRONOMY: HALE AND LUCKEY 
The need of caution is emphasized by the fact that the observed daily 
motions in latitude appear to be considerably smaller than theory would 
indicate. We have verified Kelvin's approximate expression for the 
motion of translation of vortex rings^ by means of a semi-circular vor- 
tex ring in water, with paddles which can be driven at angular velocities 
varying through a wide range. But it does not follow that the same 
expression can be applied under the very different conditions existing 
in the sun, and we are seeking further information regarding this point 
before attempting to complete the quantitative test. 
We may now inquire whether experimental methods will not throw 
some light on the difficult problem of the structure of the hydrogen 
floccuH. In 1908 we found that sim-spots are centers of vortex phenom- 
ena, plainly shown by the spectroheHograph when the sun is photo- 
graphed with the light of the red line (Ha) of hydrogen. It was subse- 
quenly proved by Evershed and St. John, however, that these are sec- 
ondary vortices, extending down through the solar atmosphere from the 
hydrogen level, where their stream-lines are visible, toward an underly- 
ing vortex which constitutes the spot and produces its magnetic field. 
Are these secondary vortices of hydrodynamic origin, as St. John has 
maintained, or are they due to the effect of the magnetic field on the 
trajectories of the electrically charged particles in the solar atmosphere? 
The latter view, impKed in Brester's aurora hypothesis of the flocculi 
and more clearly stated by Deslandres, has been fully developed mathe- 
matically by Stormer. Undoubtedly both hydrodynamic and electro- 
magnetic phenomena must influence the structure of the solar atmos- 
phere, but a satisfactory means of distinguishing between the two has 
been lacking. 
The structure of the hydrogen flocculi shown on spectroheHograph 
plates can be more or less closely imitated with the aid of smoke in a 
closed glass box above vortices in water, and photographs of the stream- 
Knes at different levels are easily obtained. The most interesting case 
we have yet observed is that produced when two paddles, rotating in 
opposite directions in water, set up a surface circulation similar to that 
in a bipolar spot group. The stream-fines in the smoke, seen when look- 
ing vertically downward toward the water, resemble those of a vortex 
ring of colored Hquid rising through water from the orifice of a tube. 
When observed from a point in the plane of the ring, the stream-fines 
in the latter case appear to be closed on the upper or advancing side, 
while the surplus fiquid on the following side moves along fines which 
are straight and axial opposite the center of the ring and become more 
and more convex on either side of this axis.^ In the smoke similar 
stream-fines appear in horizontal planes, producing the effect of a hood 
