G18 PROCEEDINGS OF THE AMERICAN ACADEMY. 



were obtained by use of the rupid methods of Golgi, and by treatment 

 with gold chloride after fixation with 10% formol. The methods of 

 Golgi and their modifications have produced such varying results in 

 the appearance of ner^'e fibers in vertebrates generall\', that a great 

 number of variations in length of time, temperature, etc., were em- 

 ployed in order to test, if possible, the reliability of the impregnations 

 obtained. Several hundred specimens were subjected to these various 

 modifications of the methods of Golgi, and enough specimens have 

 been sectioned and examined to establish some definite facts. In 

 general it was found that all animals must be cut into two or more 

 pieces, depending on their size; since whole specimens, beside im- 

 pregnating unsatisfactorily, usually show numerous artifacts. It is 

 absolutely essential that the fluids employed be used in large quanti- 

 ties, and slight warmth was found advantageous in bringing about 

 impregnation. A liberal amount of the silver nitrate should be used 

 and changed at least once. The rapid methods were by far the most 

 satisfactory, as shown by the greater number of fibers impregnated, 

 and the freedom from artifacts. Dogiel states that varicosities in 

 the motor fibers are constantly present in Golgi preparations. The 

 accompanying figures (PI. 5, Figs. 25-29; PI. 8, Figs. 43-50) show that 

 such is not the case. The difficulty in securing impregnations of 

 motor fibers after they enter the neural tube suggests that their 

 chemical composition may differ in that region. Modifications in 

 the methods employed may hereafter bring results in this case, as in 

 others. 



The impregnations with gold chloride after fixation with 10% 

 formol are comparatively free from distortion and artifacts, and are 

 probably reliable as a basis for comparison. The impregnation of 

 motor fibers and endings is somewhat uncertain with this method, 

 but when successful the preparations are very satisfactory. The 

 nerve fibers are not strikingly differentiated, but are faA'orable for 

 study on account of their accurate fixation. These preparations 

 show a tendency to fade after a consideral)le length of time. I did not 

 succeed by this method in impregnating fibers distinctly as they enter 

 the neural tube, but variations in the method may produce better 

 results. 



Strong light was particularly useful in studying the thick sections 

 necessary for tracing the nerve fibers. The shrinkage in the surround- 

 ing tissues often causes wi'inkling in fibers impregnated by the Golgi 

 method (PI. 8, Figs. 46, 47). These fibers also show a tendency to 

 break, but the artificial nature of such breaks is usually evident. 



