EXPEDITION OF THE CHALLENGER. 31 



Tesced with acid, and dried into a very light, impalpable, white powder. This, 

 when examined under the microscope, was found to consist almost entirely of 

 the frustules of diatoms, some of thera wonderfully perfect in all the details of 

 their ornament, and many of them broken up. Tlie species of diatoms entering 

 into this deposit have not yet been worked up, but they appear to be referable 

 chiefly to the genera Fragillaria, Coscinodiscus^ Chcetoceros, Asteromphalus, and 

 Bictyocha, with fragments of the separated rods of a singular silicious organ- 

 ism, with which we were unacquainted, and which made up a large proportion 

 of the finer matter of this deposit. Mixed with the diatoms there were a few 

 small GlohigerincB, some of the tests and spicules of radiolarians, and some sand- 

 pnrticles ; but these foreign bodies were in too small proportion to affect the 

 formation as consisting practically of diatoms alone. On the 4th of February, 

 in latitude 52 29' south, longitude 71 36' east, a little to the north of tlie 

 Heard Islands, the tow-net, dragging a few fathoms below the surface, came up 

 nearly filled with a pale-yellow gelatinous mass. This was found to consist en- 

 tirely of diatoms of the same species as those found at the bottom. By far the 

 most abundant was the little bundle of silicious rods, fastened together loosely 

 at one end, separating from one another at the other end, and the whole bundle 

 loosely twisted into a spindle. The rods are hollow, and contain the character- 

 istic endochrome of the Diatomacem. Like tlie Glohigerina ooze, then, which 

 it succeeds to the southward in a band apparently of no great width, the ma- 

 terials of this silicious deposit are derived entirely from the surface and inter- 

 mediate depths. It is somewhat singular that diatoms did not appear to be in 

 such large numbers on the surface over the diatom-ooze as they were a little 

 farther north. This may perhaps be accounted for by our not having struck 

 their belt of depth with the tow-net ; or it is possible that, when we found it on 

 the 11th of February, the bottom deposit was really shifted a little to the south 

 by the warm current, the excessively fine flocculent debris of the diatoms taking 

 a certain time to sink. The belt of diatom-ooze is certainly a little farther to 

 the southward in longitude 83 east, in the path of the reflux of the Agulhas 

 current, than in longitude 108 east. 



" All along the edge of the ice-pack everywhere, in fact, to the south of the 

 two stations on the 11th of February, on our southward voyage, and on the 

 3d of March, on our return, we brought up fine sand and grayish mud, with 

 small pebbles of quartz and feldspar, and small fragments of mica-slate, chlorite- 

 slate, clay-slate, gneiss, and granite. This deposit, I have no doubt, was derived 

 from the surface hke the others, but in this case by the melting of icebergs and 

 the precipitation of foreign matter contained in the ice. 



" AVe never saw any trace of gravel or sand, or any material necessarily 

 derived from land, on an iceberg. Several showed vertical or irregular fissures 

 filled with discolored ice or snow ; but, when looked at closely, the discoloration 

 proved usually to be very slight, and the effiect at a distance was usually due to 

 the foreign material filling the fissure reflecting hght less perfectly than the gen- 

 eral surface of the berg. I conceive that the upper surface of one of these great 

 tabular southern icebergs, including by far the greater part of its bulk, and cul- 

 minating in the portion exposed above the surface of the sea, was formed by the 

 piling up of successive layers of snow during the period, amounting perhaps to 

 several centuries, during which the ice-cap was slowly forcing itself over the 

 low land and out to sea over a long extent of gentle slope, until it reached a 

 depth considerably above 200 fathoms, when the lower specific weight of the ice 

 caused an upward strain which at length overcame the cohesion of the mass, 



