488 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
first, that the fines from no. 3 to no. 7 are so dim that their measurements 
must be rather uncertain, and secondly, that the error in measurement is 
large because of the width of the lines and of their blurred condition. How 
large this error may be is shown by several sets of measurements given in 
table 6, taken under very favorable conditions of fight and at different 
times, four to six measurements of each starch. 
Table 6 
Distances between Planes for Line i 
Corn starch varied from 4.99 to 5.08 Average, 5.04 
Cassava starch varied from 5.03 to 5.15 " 5.08 
Potato starch varied from 5.15 to 5.22 " 5.20 
Distances between Planes for Line 2 
Corn starch, all measurements Average, 3.85 
Cassava starch, all measurements " 3.85 
Potato starch varied from 3.85 to 3.93 " 3.88 
Compared to the figures for the salt measurements given above, these 
variations are large. But in spite of this fact, there are certain rather 
constant variations which seem to make the tables worth a few moments' 
study. 
In the first place, there is a marked similarity between the three kinds 
of starch, also between the starches and the calculated spacings for a cubic 
arrangement, as shown in table 7. 
Table 7. Comparison of d for the Starches and for a Cube Based on the Average for Line i 
Line No. 
I 
3 
4 
5 
Corn 
5.04 
3.85 
2.98 
2.64 
2.30 
Cassava 
5.08 
3-85 
2.98 
2.69 
2.36 
Potato 
5.20 
3.88 
2.98 
2.64 
2.35 
Cube based on ave. of line i 
5-II 
3.62 
2.95 
2.56 
2.29 
On the other hand, it is quite certain that there is no nice arrangement 
of the planes such as in the salt or the sugar crystals. Both the blurring 
of the lines and the irregular spacing are evidence of that. 
In column 5 of tables 3, 4, and 5 are recorded the differences between 
the actual d and the theoretical d. Line 2 shows a consistently large 
difference. Line 4 also shows a large difference, although less than line 2. 
This leads one to suspect that the arrangement is not quite cubic. The 
blurring of the lines, however, makes the error in measurement too large 
to warrant further speculation in that direction. Finer and sharper lines 
are needed before going further. 
Another point which determines certain future work is brought out by 
the no. I lines of corn starch. Here the broad line is resolved into two lines, 
I and la, while with potato and cassava starches only one broad line is 
