AND INHERITANCE IN PEDIASTRUM. 411 



cells of series II. as we pass from pole n to pole in of the colony. 

 The average values of these angles, comparing all seven colonies, 

 have an extreme range of 9° from 103° to 112°, with, as noted, an 

 average of the averages for all the colonies of 108°. In the single 

 colony, No. 55, the range of variation was 17° from 97° to 114°. 

 Colony No. 87 showed a range of variation of 30° from 90° to 120° 

 in the values of the angles odp, o^df^p^, etc. 



The average values of the five mid-basal angles of cells 7, 9, 11, 

 13 and 15 are given in Table V., section 5. We have here a case 

 in which the average values of the angles in the series of seven 

 colonies is the same as the value (133°) arbitrarily assigned in the 

 selected type diagram for the mid-basal angles of cells ii and 13, 

 and differs by seven degrees from the average value as measured 

 (140°) of these two angles in colony 55, from which the selected 

 type diagram was derived. This is the reverse of the result as we 

 have found it in the other two cases noted above and a fourth case 

 noted below in which an arbitrary value was given to angles in the 

 selected type diagram. In the three other cases the average value 

 of the angles in colony 55 as measured is nearer to the average 

 value of the corresponding angles in the series of seven colonies 

 than to the arbitrarily assigned value. As I have already pointed 

 out ('16), the angles about the point e^ in colony 55 are obviously 

 to the eye the most unsymmetrical and aberrant in the whole colony 

 and the possibility of making consistent arbitrary corrections for 

 these asymmetries and their correlated effects on the other angles 

 in the region is not very great. The values of the unpaired angle 

 oeo^ and the pair at e'^ and e'^ agree closely with those obtained from 

 the selected colony No. 55 and given in the selected type diagram. 

 The values of these mid-basal angles of the peripheral cells are 

 larger than those of the other angles of the groups about the points 

 of intersection e, e'^, e", etc., and this we may regard as a direct cor- 

 relative of the fact that the tangential diameters of the cells of 

 series II. and III. are regularly greater than their radial diameters. 

 If this tangential elongation of the cells were directly determined 

 by the fact that there are but five cells in series II. and ten cells in 

 series III., instead of six and twelve, the normal numbers for a 

 least-surface configuration in which all three angles would be 120° 



