TRANSACTIONS OF SECTION I. 979 
endings, then ‘006 second of the total time would be accounted for ; the remaining 
time would then be distributed over the other structures. 
2. The rhythm of the electrical reflex responses is a slow one in Malapterurus 
with a maximal rate of 12 per second. Superimposed on this rhythm is a rapid 
peripheral organ rhythm due to self-excitation, and in no way due to central 
nervous discharge. 
. The rate of 12 per second is not often met with, and a series of this type has 
very few members, at most two or three. The most frequent rate is one of 4 per 
second. 
The number of members of even this slower type is limited to from two to six. 
The experiments thus show that the single nerve cell discharge can occur at 
12 per second, but that it generally occurs at slower intervals, and in all cases 
rapidly fails. 
The contrast offered by these results to those of Torpedo, in which the central 
rhythm varies from 100 to 30 per second, suggests that the latter owes its rapid 
periodicity to the large number of nerve cells which innervate the Torpedo organ, 
and which are thrown into successive activity. 
SATURDAY, SEPTEMBER 19. 
The following Papers were read :— 
1. On the Principles of Microtome Construction. By Cuar.es 8. Minot, 
Professor at the Harvard Medical School, Boston, Massachusetts. 
With the advance of biology, particularly in the domains of embryology and 
cytology, we have passed during the last twenty-five years through a complete 
revolution of methods, with the result that the microtome has become as indispen- 
sable as the microscope, and hence the construction of microtomes may fittingly 
eccupy the attention of the Physiological Section of this Association. 
_ The first object of a microtome is to make sections of even and known thickness ; 
the second object is to make sections in large numbers of uniform thickness ; the 
third object is to make sections rapidly. Finally, in recent years, there has been 
a growing and justified demand for microtomes to make good sections of great 
thinness, if possible not over one five-hundredth of a millimeter or two microns 
(0:002 mm.). Now, sections which vary more than one-tenth from their supposed 
thickness, can in the case of stained animal tissues be readily recognised by the 
naked eye as uneven, hence, it is obvious that the thinner the section the less 
must be the amount of absolute error in the cutting. For example, an error of 0:002 
mm. is the maximum admissible for sections of 0:002 mm. (500 to a millimeter), 
though a much greater error would not be noticeable in sections of 0:01 mm. 
Applied to the microtome this means that a roughly made instrument is sufficient 
for thick sections, but the most perfect construction is necessary to secure a micro- 
tome for fine cutting. 
In the automatic microtome, worked by a revolving wheel, which I have 
devised, and which is now made in England, Germany and France, as well as in 
America, the attempt is made to secure mechanical perfection, and so far success- 
fully that sections of 1/300 mm. may be made with it. The microtome is, how- 
ever, adapted only to cutting objects imbedded in paraffin. The model shown is 
the latest American pattern, and has certain minor improvements which have 
increased the accuracy and precision of the instrument. 
A second microtome was also shown, which is novel in construction, and is 
suited for both paraffin and celloidine cutting. In designing this microtome 
precision was made the first object. The usual sources of error are—(1) in the 
bending of the knife; (2) the yielding of the object to be cut, chiefly because it is 
borne on an arm, which acts as a lever; (3) the ‘jumping’ of the sliding gear. 
All these defects are at their maximum in the Rivet type of microtome, of which 
