vin BLOOD-STKEAM : MOVEMENT IN VESSELS 235 



falls. This paradoxical fact was explained by Donders as the 

 result of the formation of vortices, at the base of the dilatation 

 already referred to (Chap. VII. p. 190). It is certain that the 

 vortices consume a certain amount of force. When this con- 

 sumption is considerable, pressure falls ; when, on the other hand, 

 it is inconsiderable, pressure either remains unaltered, or increases, 

 as must always occur in consequence of the diminution of velocity. 

 It is obvious that where the dilatations and constrictions of the 

 tube occur gradually instead of suddenly, the alterations of velocity 

 and pressure in its different sections must obey the same laws, with 

 the simple difference that they are produced slowly, so that their 



Fi<:. 89. Schema similar to that of preceding figure. The rigid conducting tube of equal diameter 

 is here replaced by a tube of unequal diameter in its various parts. (Rollett.) 



course is represented by a curved rather than by a straight gradient, 

 as in the preceding case. 



(d) When the tube connected with the vessel branches into two 

 or more smaller tubes so that the bed of the current is widened, 

 i.e. the sum of the sectional area of the branches presents a 

 larger diameter than the original tube, the result is complicated. 

 Since the sectional area of the current is enlarged, the sum of the 

 resistances should be diminished ; on the other hand, the branching 

 of the tube must introduce new resistances, which increase with 

 the size of the angle made by the branches. These two opposite 

 conditions are to a certain extent compensatory. Jacobson's 

 experiments (1860), however, show that the preponderating in- 

 fluence is always the widening of the bed, by which the amount of 

 fluid passing through each cross-section of the system in the 

 time-unit is increased. 



(e) If the tube after branching so as to widen the bed of the 



