108 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



graduated, if tbe volume is to be determined ; the water is emptied into a measuriug-glass of 3,000 

 cubic centimeters and you read; there is tbe difficulty, tbe water wetting tbe sides of tbe glass 

 rises on it, and one can be mistaken to tbe extent of five cubic centimeters. Another cause of 

 error, which is more important, is in tbe water wetting tlie walls of the skull itself, soaking and 

 penetrating through the internal free orifices as far as the vacuoles and sinuses. If this water 

 remained in tbe walls it would only be a half evil ; but when the skull is drained a part comes out 

 of the sinuses and vacuoles a)id is unduly measured with what comes out of the cavity proper. 



''Instead of directly seeking the volume we can proceed by weight. We weigh the skull full, 

 then the skull empty; the difference is the weight of tbe water; but comprised in it is tliat which 

 the vacuoles and sinuses contain. There is only one means of reme<lyiug this cause of error, it is 

 to saw the skull, to cover it internally with au impermeable coat and to join the two halves. This 

 (;an succeed; but in spite of the care exercised in the preparation one cannot guarantee that Ihe 

 water will not again infiltrate the walls uni)erceived. If instead of measuring tlie water as it 

 comes out of the skull it is measured as it goes in, the same causes of error persist."'* 



Elsewhere he speaks of the use of those sawed and carefully varnished skulls as terms of 

 comparison ; but even in this capacity he condemns them.t 



A careful consideration of all the literature attainable on tbe subject of water-measurements 

 led us to conclude that the experimenters had been too easily discouraged, had not sufficiently 

 contended with the difficulties which the problem presented, and that we still had a good field for 

 investigation. 



In seeking for a substance with which to coat the skull and render it water-tight, the merits 

 of fresh putty, to be removed before it hardened, were suggesteil, and it was first a]»i)lied on the 

 30th of last June. At first it gave by no means perfect results; yet it seemed to promise so much 

 that we determined to persevere in its use. It appeared from the beginning that our chief diffi- 

 culties were lack of dexterity in applying the putty and the hygroscopic nature of the osseous 

 tissue. A number of experiments were performed, the causes of error noted and means devised 

 to remedj' them. It is needless to recount all our mistakes and the various stages in the growth 

 of the system. I will, therefore, proceed at once to describe our present methods and appliances, 

 and, with these fully explained, the merits and demerits of the system can be more easily under- 

 stood. 



The following are the necessary implements and materials : 



1. Scales and weights. • 



2. An ether spray-apparatus of the pattern known as the reversible spray-apparatus with 

 revolving spray-tube. 



3. A bottle of shellac varnish, made by adding one part by measure of dry gum to nine parts 

 of strong alcohol. 



4. A roll of Seabury and Johnson's India rubber adhesive plaster. 



5. A quantity of putty ; at least'teu pounds should be kept in store. 

 C, 7, .S. Simi)le cerate, lard and linseed oil. 



9. A bread-board and rolling-pin with which to work the putty. 



10. A covered jar containing water in which to preserve the putty when it is not in use. 



11. A reservoir of water provided with India rubber tubing and stop-cock. The reservoir 

 now in use in our laboratory has a capacity of about 16 liters, is elevated 1| meters, has a tube - 

 meters long and of 13 millimeters caliber, and a stop-cock of 5 millimeters caliber; but these are 

 not essential details. 



12. An ordinary tin half gallon measure, half-covered, and provided with a s]K)ut 3 centim- 

 eters in diameter. 



13. A wide shallow pan ])rovided with a lip, for receiving the water from the skull and trans- 

 ferring it to the measuring glass. The pan we use is 36 centimeters wide and 8 centimeters deep. 



14. A metronome, set to count seconds. 



15. A measuring glass graduated for 2,000 cubic centimeters, such as that adopted by Pro 

 fessor Eanke, of Munich. 



•TOPINAUD, ioc.cit., page 592. t Id., ioc, ci«., pages 597,598. 



