BEHAVIOR AND FUNCTIONS OF THE NUCLEUS 287 



such change in the fine structure of the nucleus during clumping 

 and renodulation. According to certain of Weisz's findings (see 

 below), coalescence of the macronucleus would be necessary for 

 homogenization after individual nodes at its extremities had 

 become diverse, but this diversity could not be confirmed. It may 

 be that the nucleus in stentors coalesces into a single mass in order 

 to make possible its complete renodulation (Tartar, 1957b); for 

 in division the nucleus produces at one stroke about twice the 

 number of the original nodes which are now half-sized. This could 

 explain why coalescence is often not complete in regeneration and 

 reorganization : because the nucleus will generally return to about 

 the same size and number of nodes. 



3. Nodulation 



Following fusion into a compact mass the macronucleus extends 

 to a long and conspicuous rod or sausage shape which is then 

 renodulated, again, generally from both ends towards the middle. 

 An exception is roeselt, in which node formation proceeds only 

 from the anterior end. 



As Johnson described it, nodulation seems to involve the aggre- 

 gation of chromatic substance into serial packets separated by clear 

 nuclear material where constriction then occurs; and he also 

 remarked that the new nodes are usually " beautifully symmetrical 

 and alike in size". Rarely there is produced a forked nucleus or 

 nuclear chain with a side branch (Fig. 78B), of which Johnson 

 illustrated one case in coeruleus, also noting that Stein (1867) 

 showed a similar case in polymorphus. The same have been observed 

 in Condylostoma (Yagiu, 1952) and in Spirostomum ambiguum 

 (Bishop, 1927). These may be ectopic rejoinings of separated 

 parts of the nucleus with the main strand. 



Even after the period of renodulation, the number of nodes can 

 undergo small changes, reduction through fusion of adjacent nodes, 

 or increase either by the splitting of one node into two or by the 

 interpolation of new nodes between existing ones (Fig. 8oa). It is not 

 uncommon to find dumbbell-shaped nodes or one or more tiny 

 nodes lying between the larger. Prowazek (1904) first described 

 a step-wise increase in nodal number, and his account was 

 generally corroborated by Schwartz (1935). According to the 

 earlier investigator a node may either split in two or part of its 



