72 



SCIENCE 



[N. S. Vol. XL. No. 1019 



grams (about 50 c.c.) of ripe ovarian eggs of 

 Bana pipiens were boiled in absolute alcohol 

 and extracted with absolute ether and dried 

 at 135°. They were then powdered and boiled 

 in 200 CO. distilled water slightly acidulated 

 with acetic acid (free from salts) to coagulate 

 the proteins, and filtered. The filtrate was 

 evaporated down and both filtrate and precipi- 

 tate charred and extracted and titrated for 

 chlorides. The filtrate required 1.55 c.e. 1/10 

 normal AglSTOj, whereas the precipitate re- 

 quired but .2 c.c, which might be due to the 

 small amount of filtrate held in the precipitate. 

 It thus appears that very little if any salt was 

 adsorbed. If all this chloride is NaCl it would 

 make a .00766 molecular solution of the same 

 volume as the egg. However, the osmotic 

 pressure of the ovarian egg corresponds to 

 that of a .166 normal NaCl solution. If this 

 osmotic pressure is due chiefly to NaCl it 

 must be confined to the watery phase which 

 must equal .0456 or about 1/20 of the volume 

 of the egg. 



I found that frog's eggs lose NaOl continu- 

 ously during their development in distilled 

 water, hence they must be permeable to NaOl 

 for some time after fertilization. This is in 

 harmony with the fact that pure NaCl solu- 

 tions are not so toxic to the frog's egg as to 

 the eggs of many other animals. I found that 

 those salt solutions which were toxic to fish 

 eggs increased the permeability, but the fer- 

 tilized frog's egg is already permeable.^" 

 Some of the older work on the effect of pure 

 NaCl on the frog's egg might be objected to on 

 the ground that the NaCl solution became 

 contaminated by Ca contained in the egg 

 jelly. Therefore I made a series of experi- 

 ments in which small numbers of frog's eggs 

 were washed for an hour in several liters 

 of distiUed water, and placed in several 

 liters of pure NaNOj solution. Very dilute 

 solutions were non-toxic. One tenth molecular 

 solutions showed a toxic effect in 48 hours, 

 but this may have been due to osmotic pres- 

 sure, since the addition of 1.6 c.c of a molec- 

 ular CaCl, solution to the liter did not decrease 



10 Although it is more permeable to water than 

 to salts. 



the toxicity. The toxicity of all salts is not 

 due entirely to osmotic pressure, since I found 

 lithium salts to be slightly more toxic than 

 sodium, salts of same osmotic pressure. 



All of the abnormalities in the lesser toxic 

 salt solutions which I have observed or found 

 in the literature, are characterized by a retar- 

 dation or failure of the white pole to segment. 

 This is also true of abnormalities produced by 

 centrifugal force or other mechanical agents 

 applied to the unsegmented egg. This unseg- 

 mented white pole prevents or retards the 

 downgrowth of the black cell layer, and in 

 extreme cases leads to the so-called " lithium 

 larvse." These embryos may regenerate and 

 become normal tadpoles. The more toxic solu- 

 tions prevent segmentation of the white pole 

 and cause swelling of serous cavities (peri- 

 cardium) and a separation or loosening up of 

 the black cells, accompanied by death of some 

 of these cells (a condition called by Eoux 

 "framboisea"). This condition (also seen 

 in fish embryos) occurs after the frog's embryo 

 has partially regained its semipermeability, 

 and may be due to an abnormal increase in 

 permeability by the salt solution. 



J. F. McClendon 



Physiological Labobatoey, 

 Medical School, 

 University op Minnesota, 

 June 1, 1914 



TEE AMERICAN CHEMICAL SOCIETY. Ill 



DIVISION OP PHYSICAL AND INORGANIC CHEMISTBY 



Gr. A. Hulett, Chairman 



R. C. Wells, Secretary 



Ea.pi4 Detection of Arsenic in Poison Cases iy the 



Marsh Test: James E. Withrow. 



It seems to have been the experience for a long 

 time that the number of cases where arsenic is the 

 poison used exceeds that of all other poisons com- 

 bined. Certain and rapid detection is therefore 

 a matter of much moment. Any effort to make 

 old methods more certain and to eliminate possi- 

 bility of error by contamination or to abbreviate, 

 thus reducing opportunity for loss, axe desirable. 

 The Berzelius-Liebig modification of the Marsh 

 test (1836) has long enjoyed confidence as one of 

 most satisfactory tests. It requires for universal 

 certainty of results the elimination of organic 



