mold aggregation. In this model, the attractant 

 (acrasin) produced by one cell causes adjacent 

 cells to migrate toward the source of acrasin 

 and to, in turn, produce acrasin. The acrasin 

 produced by the affected cells stimulates other 

 cells to do likewise resulting in the "relay" 

 and "amplification" of the aggregation message. 

 Contact following is a term used by Shaffer (3) 

 to indicate that cells in a stream adhere and 

 follow one another. Like circus elephants, the 

 cells in a row follow the lead cell. How informa- 

 tion regarding speed and direction of movement 

 is relayed from cell to cell is not known. 



Our attention, to this point, has been focused 

 on those interactions which take place between 

 cells. The final item in Table I refers to an inter- 

 action at the multicellular level. This interaction 

 is manifest in the disposition of centers (cen- 

 ters of aggregation) with respect to one another. 

 More precisely, certain evidence indicates that 

 the presence of one center may dictate whether 

 a second center can form within the immediate 

 area. 



Before discussing this phenomenon, are 

 there any questions? 



GREGG: Arnold, would you care to com- 

 ment on the fact that you can get aggregations 

 within a mass of bacteria on occasion? 



KAHN: I haven't seen this occur myself, 

 but I can think of a possible explanation. If the 

 bacterial attractant is short-lived (acrasin is 

 short-lived under normal conditions), then a 

 point may be reached where it would no longer 

 compete with acrasin and aggregation could 

 proceed. 



GRUN: If you take an amoeba from a 

 colony which is aggregating and if you put it 

 into the middle of a colony which is vegetative, 

 does it pass the message to the others? 



KAHN: No. However, Sussman has shown 

 the aggregative phase amoebae can stimulate 

 aggregation in developmentally younger cells. 



If there are no further questions, I should 

 like now to return to the last item in the table. 

 My interest in this problem arose as the result 

 of several investigations carried out by Bonner 

 and co-workers (4-6). Their studies indicate 

 that the orientation of fruting bodies and the 

 number of aggregates formed per unit area of 

 substrate may be under the control of a factor 

 present in the gaseous phaseof the environment. 

 They termed thisfactor the "spacing substance." 



I began my study in the hope of answering 

 two questions. First, does the spacing of aggre- 

 gates occur in Poly sphondy Hum pallidum? Sec- 

 ond., if such spacing does occur, is it the result 



^ 



4< ^ ^ 







sV Vv W vJ< 



A 

 clustered 



random 



Fig. 2, 



C 



spaced 



Three types of possible spatial distribution of aggrega- 

 tion centers. A) Clustered, centers appearing In groups; 

 B) random, centers distributed as expected on the basis 

 of chance; C) spaced, centers placed at equal distances 

 from one another. The density of centers is the same in 

 all three examples. 



of a spacing substance present in the gaseous 

 phase of the environment? Previous work with 

 Polys phondy Hum indicated that this species was 

 responsive to those factors (charcoal, mineral 

 oil) used by Bonner to reduce or eliminate the 

 spacing substance. 



Spacing may be defined as the distribution 

 of centers of aggregation on a substrate. A 

 "spaced" distribution is one in which the cen- 

 ters tend to form at equal distances from one 

 another (Fig. 2C). ^ A "clustered" distribution, 

 on the other hand, is one in which the centers 

 tend to appear in groups (Fig. 2A). The method 

 of Clark and Evans was used to determine the 

 distribution of centers. This method consists of 

 calculating the nearest neighbor distance ex- 

 pected if the distribution is at random and com- 

 paring this value with one derived by actual 

 measurement. If the distribution of centers is 

 random, the ratio of observed to expected is 

 unity. If the distribution is spaced, values 

 greater than one are derived; if clustered, the 

 values are less than one. Figure 3 is a graphic 

 illustration of the relation between nearest 

 neighbor distance and the density of aggregation 

 centers. Note that deviations to the right of the 

 curve indicate a spaced distribution while devia- 

 tions to the left indicate clustering. 



In these experiments, the cells were pre- 

 grown in liquid culture, washed free of residual 

 bacteria by differential centrifugation, sus- 

 pended in a saline solution, and dispensed in 



1 Figures 2-6 are sketches of data which will appear in 

 Developmental Biology, 1966. 



125 



