3. From what we observe in regard to the _larger_ bodies of the universe--a clustering tendency being everywhere apparent,--it seems highly improbable that cometic matter should be uniformly distributed in the sidereal spaces. We would expect, on the contrary, to find it collected in groups or clusters. This view is also in remarkable harmony with the facts of observation. In 150 years, from 1600 to 1750, 16 comets were visible to the naked eye; of which 8 appeared in the 25 years from 1664 to 1689. Again, during 60 years, from 1750 to 1810, only 5 comets were visible to the naked eye, while in the next 50 years there were double that number. The probable cause of such variations is sufficiently obvious. As the sun in its progressive motion approaches a cometary group, the latter is drawn toward the centre of our system; the nearer members with greater velocity than the more remote. Those of the same cluster would enter the solar domain at periods not very distant from each other; the forms of their orbits depending upon their original relative positions with reference to the sun's course, and also on planetary perturbations. It is evident also that the passage of the solar system through a region of space comparatively destitute of cometic clusters would be indicated by a corresponding paucity of comets. 4. The line of apsides of a large proportion of comets will be approximately coincident with the solar orbit. The point towards which the sun is moving is in longitude about 260°. The quadrants bisected by this point and that directly opposite extend from 215° to 305°, and from 35° to 125°. The number of cometary perihelia found in these quadrants up to July, 1868 (periodic comets being counted but once) was 159, or 62 per cent.; in the other two quadrants, 98, or 38 per cent. This tendency of the perihelia to crowd together in two opposite regions has been noticed by different writers. 5. Comets whose positions before entering our system were very remote from the solar orbit must have _overtaken_ the sun in its progressive motion; hence their perihelia must fall, for the most part, in the vicinity of the point towards which the sun is moving; and they must in general have very small perihelion distances. Now, what are the observed facts in regard to the longitudes of the perihelia of the comets which have approached within the least distance of the sun's surface? But three have had a perihelion distance less than 0.01. _All_ these, it will be seen by the following table, have their perihelia in close proximity to the point referred to: =I.= COMETS WHOSE PERIHELION DISTANCES ARE LESS THAN 0.01. Perihelion Passage. Per. Dist. Long. of Per. 1. 1668, Feb. 28_d._ 13_h._ 0.0047 277° 2´ 2. 1680, Dec. 17 23 0.0062 262 49 3. 1843, Feb. 27 9 0.0055 278 39 In Table II. all but the last have their perihelia in the same quadrant.