300 A. FRANKLIN SHULL 



is relatively large and stains deeply, the drawings could be made 

 with a micro-projection lantern. Because it is a syncytium, it 

 was necessary to draw sections of the entire organ, not individual 

 cells. The nuclei are not appreciably longer in one direction 

 than another, so that sections in all planes are equally good. The 

 only requirement of a section to be drawn was that the boundaries 

 of both the organ as a whole and of its nuclei be sharp and dis- 

 tinct. This requirement could be met only in case the section 

 passed near the middle of the nuclei. The first two or three 

 sections through a given nucleus were invariably rejected be- 

 cause they were too nearly tangential, while frequently more 

 than three were rejected. 



Although the yolk gland usually contains eight nuclei, and as 

 many as five (rarely more) nuclei might appear in a single sec- 

 tion, it was seldom that a section passing through many nuclei 

 passed near the middle of all of them. The drawings were 

 necessarily made, therefore, mostly from sections passing through 

 one, two, or three nuclei, the smaller numbers more frequently 

 than the larger. 



Measurement of the drawings. All measurements were made by 

 means of a polar planimeter. Since absolute measurements 

 were not desired, no tests of the absolute accuracy of the pla- 

 nimeter have been made. The value of the planimeter in this 

 w^ork depends on the uniformity of its measurements. Several 

 tests of the instrument were made by measuring a single figure a 

 number of times, always in the hands of the same operator. 

 One such test resulted as follows: 



1 time, 4.55 square inches 



3 times, 4.56 square inches 



4 times, 4.57 square inches 

 3 times, 4.58 square inches 

 3 times, 4.59 square inches 

 1 time, 4.60 square inches 



While there is some irregularity, there are no large deviations 

 from the mean. A single measurement of each figure was there- 

 fore regarded as adequate. 



