150 
sublimate in 5 °/, glacial acetic acid. After hardening, dehydrating, 
and embedding in paraffin through cedar oil, the sections were cut 
after two months had elapsed. They were then passed through iodine 
and potassium iodide and stained in iron haematoxylin. In these sections 
structures were found which resembled closely the so-called centro- 
spheres and centrosomes (see photographs). 
Fig. 1. Fig, 2. 
The central disk was somewhat darker than the ground color of 
the cytoplasm and was also free from the granules. It was usually 
somewhat larger than the nucleus and generally regular in form. A 
black spot, or corpuscle, appeared in its centre, and, by focussing, a 
radial arrangement of the disk was visible. These disks always occupied 
the same relative position in all the cells — namely between the pig- 
ment mass and nucleus and generally near the exit of the neurite. 
Also, there was a very distinct concentric arrangement of the granules 
of the cytoplasm around this disk. 
The spinal ganglia of other dogs were prepared in a similar 
manner and the disks, with their accompanying structures, were found 
again. It was only after a careful examination of the sections in the 
different stages of preparation that the artifact was revealed. 
That this disk, and its contained structures, are artifacts I consider 
proved by the following facts: 
Before placing the sections in the iodine mixture, crystals of sub- 
limate, corresponding in position, form and size to the above mentioned 
disk, were found in the cells. I have made this observation in about 
twenty-five different cells; observing and marking the positions of the 
crystals in the cells; then dissolving these crystals out, staining the 
sections, and observing that the structures mentioned were in the same 
places that the crystals had occupied in the cells before treatment. 
