CYTOPLASMIC CONSTITUENTS OF NERVE CELLS 399 



Embryos imbedded in paraffin were cut in serial sections by 

 means of a Minot rotatory microtome. The great majority of 

 the series were cut transversely; but others were made in coronal 

 and sagittal planes so that the one-sided conceptions so likely to 

 arise from the perusal of sections cut in a single direction, have 

 been eliminated. All the material imbedded in paraffin was cut 

 into sections 4 /z thick. 



The celloidin method of imbedding was employed, in the case 

 of several embryos, on account of the objections raised by cer- 

 tain authors (Cajal '07, p. 172, et al.) to paraffin. Complete 

 series were made about 7 m thick by the technique advocated by 

 Maxunow ('09, p. 184). 



This investigation has been controlled as follows: 



(1) The entire nervous system has been studied; but detailed 

 comparisons have only been made between the results of differ- 

 ent methods in the neural tube and the neural crest on the right 

 side in the region of the sixth somite of embryos of the same 

 degree of differentiation. 



(2) The commoner artefacts due to fixation are shrinkage, dis- 

 tortion of form relations, and destruction, solution or change in 

 the nucleus and cytoplasmic constituents. Shrinkage is de- 

 pendent, to some degree, upon the water content, and since this 

 probably varies in different regions we must look for unequal con- 

 traction. The diminution in the longitudinal axis was measured 

 in the case of the mitochondrial and neurofibrillar fixatives; in 

 the former it amounted to 23 per cent, in the latter to 19 per 

 cent; (compare the size of the nuclei represented on plates 3 

 and 4) . The finer form relations, as seen after fixations in vogue 

 for mitochondria and neurofibrils, were controlled by comparison 

 with the results obtained after a large number of fixations, es- 

 pecially adapted for their accurate delineation, and with living 

 tissue teased out in warm physiological salt solution. It ap- 

 pears that the shape of the cells is but little altered by the fixa- 

 tions, and I am inclined to believe that, in the chick, the difficulty 

 encountered in the observation of cell boundaries is in some meas- 

 ure due to shrinkage. Moreover, Harrison ('10, p. 793) states 

 that the cells in the neural tube of frog embryos are perfectly 



