64 
whose body vertebre alone may vary as much as one hundred. 
When investigating this character one should record the 
number of body and caudal vertebre separately and obtain 
their sum. 
VARIATION IN SUM OF GASTROSTEGES AND UROSTEGES 
The number of verterbre in terms of gastrosteges and 
urosteges is subject to a fairly wide range, being from 170 to 
226, a difference of 56. Specimens from Siberia possess the 
highest, those from the Yangtse Valley the lowest, and those 
from Korea and Japan the intermediate counts. 
Examples. Minimum. Maximum. Average. 
152 6 
Ran ren ot eval thor cetehte ccetelshaiers (ola edeleleletere 170 190 
Range in Mainland specimens.......... 109 170 193 
CentralIS Dera ais cies iofeeskeletel crereterstel« 29 195 214 
BIASCSTM OT GEILE ialarsie'syslate/eiete\elsielsis)olsi= 27 191 195 
Gove, Ane oso c ocapuubdcmpe uNoDeooUNO 24 175 187 
(Qi, 0h Gk Gen ow cepsadue donodppaedonod 29 170 178 
Range in Island specimens ............ 43 179 187 
Hokkaido 6 183 188 
Fond omen <sec 23 179 191 
Shikoku 2 187 187 
Hachijosh 2 185 185 
Awaji 1 190 190 
Kiushu 1 184 184 
Ikishima 5 182 185 
Tsushima 3 182 188 
The range of variation in the mainland examples being 
from 170 to 226 and of the Island examples from 179 to 200, 
it is evident that the range of the mainland specimens widely 
overlaps that of the island specimens. 
VARIATION IN GASTROSTEGES. 
Every effort should be made to have the gastrostege 
count an index to the number of the dorsal vertebre. To 
further this aim it is necessary to ascertain which shield 
corresponds with the first link in the vertebral column. A few 
dissections have shown the most approximate estimate that 
can be made without actually opening the neck is to reckon as 
the first gastrostege the first shield that is of the same width 
as the entire series; it is usually further characterized by 
being colored like the rest of the gastrosteges and not white 
as are the throat shields. 
The shields between the anterior pair of geneials and the 
first gastrostege are referred to as gulars; they are of two 
types, paired and azygos. Note should be made of their num- 
ber in the record of each specimen, as this will tend to fix the 
first gastrostege, this being the shield from which are started 
many of the counts. 
If the locality of each specimen be plotted on squared 
paper, due regard being taken of the gastrostege and urostege 
