210 THE AMERICAN MONTHLY [November, 
In view of these observations, it would appear that we must accept an 
amendment to the current statement of the facts as to the origin of the Pro- 
sobranch foot to the effect that in some cases (so far as at present known two 
cases) the organ is at first a paired structure like the beginnings of locomotor 
appendages in arthropods and vertebrates, and only later in the ontogeny 
median in position. 
Further, we may open the question whether these facts may be taken to 
have phylogenetic significance or to be a falsification of the phylogenetic 
record. The latter view would seem improbable, because the foot in gas- 
teropods is of no use in the early larval life, the velum being the organ of 
locomotion. 
It would be equally difficult to see how the paired condition of the foot 
could have been acquired secondarily from an ancestral form with a single 
median one, both because of its probable rarity in the group and because of 
the absence of function of the foot in early life. On the other hand, the sup- 
position that the gasteropod foot originated as a median structure from the 
ventral creeping surface of a Vermian ancestor would seem natural and easy, 
being in accord with the facts of anatomy and the few hitherto published facts 
of embryology. In thus seeking a meaning for this appearance in these two 
forms, which deviates from the position commonly held, I have no desire to 
attempt to force a few facts to support a superstructure of speculation but 
rather to propose the question for future determination when a larger number 
of forms shall have been studied. 
Report on methods of stating the results of water analyses. 
By G. C. CALDWELL, 
ITHACA, N.Y. 
At the Buffalo meeting of the American Association for the Advancement 
of Science a paper on this subject was read before the Chemical Section by 
Prof. W. H. Seaman, embodying the report presented to the Washington 
Chemical Society by a committee of that society. Some of the results ar- 
rived at by that committee, and their recommendations, are given in the 
following extracts from that report. 
After giving a ‘ list of forty-two methods of statement or expression, based 
on an inspection of about a thousand analyses,’ and noting that ‘in some 
places three scales were found in the same table, a part expressed as grains 
per gallon, another part as milligrammes per litre, and the hardness in Clark’s 
degrees,’ the report goes on to say :— : 
‘These can nearly all be reduced to one or another of four common 
methods of expression as detailed by Nichols, viz: 
ist. Grains per English or imperial gallon (277 cubic inches or ro Ibs.— 
70,000 grains of pure water). 
2d. In grains to the U. S. or wine gallon (231 cubic inches—58,372 + 
grains of pure water). 
3d. On a decimal basis as parts per 100, 1,000,1,000,000. This is gen- 
erally used in France and Germany, also in the Reports of the Rivers 
Pollution Committee of Great Britain, and in this country in reports of the 
National Board of Health and of many State Boards of Health. 
4th. As so many milligrammes to the litre. (This would be the same as 
parts in 1,000,000 if the water had the same density as pure water). 
‘ After careful consideration the committee would recommend the follow- 
ing conclusions, viz :— 
ist. That water analyses be uniformly reported in parts per million or mil- 
