### Table 2 Mean statistic of inter-cell control group

"... In PAGE 6: ... Ideally, we should be able to assume that the local expansion rate areas of all control cells, when differenced against each other should have mean zero. Indeed, in Table2 , the control spontaneous case has a small mean for embedding dimensions 6 and 7, but the control driven cases do not. The implication may be that there is a lack of local determinism in the driven cases, or more likely, noise is dominating the dynamics in between spikes.... ..."

### Table 7 : Variation from expected inter-cell arrival times

"... In PAGE 12: ... At the lower bit rate the time it take to handle a signal is subtracted from the end delay. The data shown in Table7 backs this statement up. Figure 7: Application Latency Tolerance and Network Product Latencies 11 Conclusions This paper provides three useful contributions.... ..."

### Table 4: Expected and Observed Values of Inter- cell Arrival Times

"... In PAGE 10: ...1 Switch Timing Results and Analysis The experiments carried out using the HP OC3Port analyzer confirmed that even though the inter-cell arrivals times vary their mean value remains close to the expected value of the inter-cell arrival time. These results are summarized in Table4 . The 95% confidence interval on all values is in the range of 0.... ..."

### Table 4: Expected and Observed Values of Inter- cell Arrival Times

"... In PAGE 10: ...1 Switch Timing Results and Analysis The experiments carried out using the HP OC3Port analyzer confirmed that even though the inter-cell arrivals times vary their mean value remains close to the expected value of the inter-cell arrival time. These results are summarized in Table4 . The 95% confidence interval on all values is in the range of 0.... ..."

### Table 7: Time-Varying EM. 14-node topology. Bad prior.

"... In PAGE 7: ... We retested the EM method using a window size of 10, to take advantage of multiple measurement intervals. Table7 shows the results. The constant case is not shown because it is not affected by incorporating multi- ple measurement intervals.... ..."

### Table 7: Time-Varying EM. 14-node topology. Bad prior.

"... In PAGE 7: ... We retested the EM method using a window size of 10, to take advantage of multiple measurement intervals. Table7 shows the results. The constant case is not shown because it is not affected by incorporating multi- ple measurement intervals.... ..."

### Table 7: Comparison of binary method dispatch times for varying degrees of modularity

"... In PAGE 7: ...able 6: Comparison of augmenting method dispatch times for varying degrees of modularity .... 89 Table7... In PAGE 96: ... This difference must be due to the additional complexity created by copying state between visitors and retrieving results via a separate method invocation. Table7 compares the performance of the multiply operations implemented using MultiJava multi- methods and using double-dispatch. The table shows that multiple dispatch is substantially slower than double-dispatching.... ..."

### Table 6: Comparison of augmenting method dispatch times for varying degrees of modularity

"... In PAGE 7: ...able5:Comparisonofdispatchtimesformultiplyoperation.............................. 88 Table6 : Comparison of augmenting method dispatch times for varying degrees of modularity .... In PAGE 95: ...) To measure the dispatch speed we instantiated one real, one integer, and one rational and invoked the multiply operation 1,000,000 times on each possible combination (for a total of 9,000,000 invoca- tions.) This test was repeated for both implementations; the results appear in Table6 . The table shows that the multiple dispatch and typecases approach yield the same performance.... In PAGE 96: ... This is an apples-to-oranges comparison, but it provides valuable information, particularly for choosing imple- mentation strategies when one knows that the code will not have to be extended.37 Table6 compares the performance of the pretty-print operation for trees implemented in regular Java code, using the visitor pattern, using extensible visitor, and using an external generic function in MultiJava. The intent of these tests is to compare the cost of modularity for various partial solutions to the augmenting method problem.... ..."

### Table 4 Cycle time varying the number of sections with two trains

"... In PAGE 22: ... The degree of interference seems to depend on the number of sections of a circuit or, conversely, on the number of trains in a given circuit. Thus we com- puted the cycle time at varying number of sections with two trains #28 Table4 #29, and at varying number of trains with 11 sections #28Table 5#29.... In PAGE 23: ... Table4 reports the optimal values of the cycle time; i.e.... ..."

### Table 5 Cycle time varying the number of trains with 11 sections

"... In PAGE 22: ... The degree of interference seems to depend on the number of sections of a circuit or, conversely, on the number of trains in a given circuit. Thus we com- puted the cycle time at varying number of sections with two trains #28Table 4#29, and at varying number of trains with 11 sections #28 Table5 #29.... In PAGE 23: ...98#25 #28for 12#29. In Table5 , it should be noted that, with increasing number of trains, the dependency makes the cycle time increase over linearly.... ..."