A. B. Lyons — Composition of Hawaiian Soils. 429 



the time covered by the history of these rocks, I have made 

 these jointly a basis of comparison, to bring out the nature of 

 the changes that have taken place in the weathering of the 

 lava and its conversion into soil. While generalizations can- 

 not be based too confidently on the results of so small a num- 

 ber of analyses, I believe that the data are sufficient to yield 

 us some quite definite and positive conclusions. The validity 

 of the general conclusions is confirmed by the specific example 

 given in the last two columns, where analysis has been made of 

 the identical rock from which the sample of soil has been 

 derived. 



• We see then that, in the incipient weathering of lava that 

 we find in volcanic tufa (which consists of lava fragments jt?^w5 

 calcium and magnesium carbonates, and some clay and sea 

 sand), that there has been a notable reduction of silica (17 per 

 cent), titanic acid (practically all), manganese (44 per cent), 

 soda (55 per cent), potash (19 per cent), copper (60 per cent), 

 while calcium remains unchanged and magnesium seems actu- 

 ally to have been increased. [This may have come from the 

 sea water.] There has been, moreover, an oxidation of three- 

 fourths of the ferrous iron. 



In the ''rotted" lava, there has been a loss of more than 80 

 per cent of the silica, while titanic acid has been but slightly 

 reduced. [If we omit from the comparison sample J^o. 2 of 

 the rotted lavas of Table III, the loss of silica would be less 

 considerable, that of titanic acid greater.] There has been a 

 loss of manganese (72 per cent), magnesia (96 per cent), soda 

 (95 per cent), copper (30 per cent), and phosphoric acid (74 

 per cent), while lime and potash have almost wholly disap- 

 peared. Sulphates, on the other hand, appear to have in- 

 creased, but possibly this would not prove generally true. 



In the soil, we find a loss of more than half the silica, 77 

 per cent of the manganese, 93 per cent of the lime, 91 per 

 cent of the magnesia, and about 50 per cent of the phosphoric 

 acid. Sodium and potassium are largely diminished, but not 

 to the extent indicated in the table, since these bases are only 

 partially extracted by hydrochloric acid. In soils like that of 

 Punahou, formed from lapilli in which the original material 

 is as yet only partially disintegrated, there may be an actual 

 increase in the proportion of phosphoric acid. In the Puna- 

 hou soil the reduction of silica amounts to only 32 per cent, 

 of manganese to 61 per cent, hme to 72 per cent, magnesia to 

 78 per cent and sulphuric acid to 55 per cent. 



It is especially interesting to note that while the rotted lava 

 has lost positively nearly all its calcium and potassium, the 

 soil retains a considerable proportion of both these elements, 

 a result to be attributed, in part at least, to the influence of 

 plants and, ii^ the case of calcium, of molluscous animals. 



