874 



SCIENCE. 



[ti.ti. Vol. XV. No. 387. 



"In tlie magnetic observatory at Agincourt, 

 nine miles from Toronto, the bifilar magnet 

 began to swing at 2 h. 35.0 m., Greenwich 

 mean time, and was at its maximum swing at 

 2 h. 44.4 m. Facsimiles of the records at 

 Toronto, Victoria and Agincourt are given in 

 the accompanying figures, Nos. 1, 2, 3, respect- 

 ively." 



C. A. 



Washington, 

 May 7, 1902. 



canaliculi in specimens injected in the labora- 

 tory of the late Professor Rutherford, of 

 Edinburgh. 



Some time ago the writer was at work on the 

 vascular supply of the chief organs of verte- 

 brates. In the course of these studies, many, 

 probably fifty, injections were made with 

 gelatine-carmine, well acidulated, of the livers 

 of the rabbit, rat, cat and other mammals. 

 The injection was made through the portal 

 vein in every instance. The most painstaking 



■'"r II (I III - y...^.^.^..^ r- 



Fig. 2. Record of Milne Seismograph, Victoria, April 18, 1902. 



i^i*** 



,«»jO ((•/».. 





Fig. 3. Record of the Bifilar and other Magnets, Agincourt, April 18, 1902. 



SHORTER ARTICLES. 



INTRACELLULAR CANALICULI OF THE LIVER. 



The cells of the liver in one of the higher 

 vertebrates are characterized by activities at 

 once numerous and diverse. It should not be 

 surprising therefore if a specialized mode of 

 communication were found between the he- 

 patic cells and the blood capillaries with 

 which they are related. Such a connection 

 has, in fact, been predicted. 



In a series of studies published from 

 Cracow, Browicz has indicated grounds for 

 believing that minute channels exist in the 

 cells of the liver, and that these are continua- 

 tions of the vascular system. More recently, 

 Schaefer* has noted the presence of such 



* Schaefer, E. A., Anatomisoher Anseiger, 1902, 

 Bd. 21, S. 18-20. 



precautions were always taken to secure the 

 conditions which experience had shown to be 

 necessary for proper injection, and some of 

 the results were as nearly perfect as could be. 

 In sections made from successfully injected 

 livers, the network of lobular capillaries is 

 uniformly filled, there is no indication of ex- 

 travasation, and the hepatic cells show no sign 

 of distortion. In such specimens, exceedingly 

 tenuous canaliculi may be seen within the 

 cells, filled with the red injecting-mass, 

 branching more or less, and anastomosing 

 with each other. Under a high magnification, 

 the blood capillaries do not present perfectly 

 smooth walls, but exhibit minute, spine-like 

 elevations at intervals. The connection be- 

 tween the intracellular canaliculi and the 

 outpushings of the capillary wall may occa- 



