1244 
7. A second series of observations has been made for points of 
the tube at distances between 20 and 30 em from the point V. By 
using a stronger source of light, | could now make observations 
in a plane perpendicular to the axis of rotation of the mirror so 
that the correction of «° to « can be omitted. The “effective” length 
/ was 31.3 em and v was 374.0. 
In the same way as above the measurements have been collected 
in this table. 
V T Vo a Un |P==L,/ Um 
1165 1067 315.2 45.4 | 368.8 0.855 
1757 1625 B1222) Lt 368.8 0.846 
1075 979 317.0 45.2 | 371.4 0.854 
1497 1383 312.5 45.3 370.1 0.844 
886 814 314.3 45.2 371.4 0.846 
762 696 | 316.1 45.3 370.1 0.854 
0.850 
8. Finally, a few observations bave been made in the initial part 
of the tube, in the neighbourheod of /. 
With the effective length / = 31.8 em and v= 380,0 em, the 
measurements gave the following result: 
| | | | 
T 
i dP v | a | Um PD Vm 
1346 | 1267 306.7 | 45.8 | 369.5 | 0.830 
| | 
1038 956 313.5 | 44.6 385.3 | 0.814 
| | | | 
1358 |) 1261) 311059") 4525 | 373.4 | 0.833 
0.826 
9. As is shown bv the experiment of § 3 the motion near / is 
more complicated than that further away in the tube, so that the 
result of § 8 cannot directly be compared with those of $$ 6 and 7. 
The image observed in the rotating mirror is also less simple in 
the experiments of § 8 than in the other series of observations. 
In the latter experiments a distinct principal direction of the 
stream-lines can be indicated; in the experiments of § 8 this is only 
