V. Results

Figure 5a is a plot of SB _IR versus L_IR/L_B for all the Sbc-Sdm galaxies in the supernovae sample which have had spectrally typed SNII or SNIb/Ic. Figure 5b is the corresponding plot for Sbc-Sdm galaxies in the supernovae sample which have had SNIa/I*.

Note: We have included the nine galaxies without HI indices in theses two plots because it is very unlikely that they are HI deficient since only a small fraction of the sample spirals outside the Virgo cluster are HI gas depleted. The data points for these nine galaxies are highlighted using darker symbols. Galaxies with high inclinations (Log(R ₂₅) > 0.45) have been highlighted with a horizontal bar.

As expected, the SB_IR for both galaxies types systematically increases with increasing L_IR/L_B. However, there is a clear difference between the distributions of the data points for the two galaxy groups. Figures 5a and 5b show that while the data points for SNII galaxies extend over the full range in S_IR and L_IR/L_B, there are very few SNIa/I* galaxies with data points above and to the right of the point (SB _IR =35.0, L_IR/L_B=0.60) (Two SNIa/I* galaxies versus 10 SNII galaxies). Note: SB_IR is in units of 10 ⁶ L_O Kpc^-2.

One way to gauge whether or not there is real lack of SNIa/I* galaxies with high current to integrated star formation rates is to compare the SN sample galaxies with the parent population from which they are drawn.

                                                     Table 4: The Base Sample

Hubble Type	Plotted Points	HI deficient	No mfir and/or BoT	TOTAL
Sbc	125	16	34	175
Sc	172	8	43	123
Scd-Sdm	159	4	102	265
TOTAL	456	28	179	663

The SN sample galaxies are drawn from a parent sample that consists of all of the Sbc-Sdm spirals in RC3 catalogue with Vo (3K) < 3000 km/sec that are also members of the NGC and Index Catalogue (de Vaucouleurs et al. 1991). There are 663 spiral galaxies that meet these criteria. Of these 663 galaxies, 179 have no far infrared magnitude (m_fir) listed and/or no total blue apparent magnitude (B^o_T) (de Vaucouleurs et al. 1991). Hence, no SB_IR and/or L_IR/L_B values can be calculated for these galaxies. In addition, 28 galaxies are HI deficient, leaving a total of 456 galaxies from which the SN sample is drawn. Here after, we will refer to these 456 galaxies as the Base sample. Table 4 shows the break down of Base sample by Hubble type.

Figure 6 is a plot of SB _IR versus L_IR/L_B for all the galaxies in the Base sample. The same general trends evident in Figure 5a are also observed in this plot, with data points extending all the way up to the region near the point (SB _IR =90.0, L_IR/L_B=1.20).

G iven that the galaxies in the SN sample are drawn from parent population plotted in figure 6, we should be able to use data point distribution in this figure to determine how rapidly the data point density falls off as you move towards higher values of SB_IR.

Figure 7a shows the percentage of Base sample galaxies that have a SB_IR greater than or equal to a given SB_IR. Curves are shown for each of the main Hubble types (Sb, Sc, and Scd-Sdm) as well as the whole Base sample (Base). It is clear from figure 7a that curve for Scd-dm galaxies falls off much faster than that for Sc galaxies , which in turn falls off much faster than that for Sbc galaxies. Not surprisingly, the curve for the whole Base sample is very similar to that for Sc galaxies, since there is roughly the same number of Sbc galaxies in the Base sample (125) as there are Scd-dm galaxies (159).

If we make naive assumption that supernovae show no preference for any particular galaxy grouping within the Base sample [other than a preference related to Hubble type], then the expected percentage of SN sample galaxies that have a SB _IR greater than or equal to a given SB _IR can be constructed by taking the weighted mean of the curves for the three Hubble groups displayed in figure 7a. [Note: this is equivalent to assuming that SN sample galaxies of a give Hubble type are selected at random from the parent population for that Hubble type.] The weights used in calculating the mean would be given by the ratio of the number of Sbc : Sc : Scd-Sdm galaxies in the SN sample expressed as a decimal percentage of the sample total. Table 5 shows the actual weights used.

                                                             Table 5

SN Galaxy Type	Sbc	Sc	Scd-Sdm	TOTAL
SN II	12	33	16	61	Number
	0.197	0.541	0.262	1.00	Decimal %
SN Ia	12	5	5	22	Number
	0.545	0.227	0.227	1.00	Decimal %

In reality, historical supernova searches show a strong biased towards galaxies with bright blue apparent magnitudes (B_T ⁰ ) (van den Bergh 1987). However, figure 7a shows that if we exclude the faintest galaxies from the Base sample (i.e. we only include galaxies with B_T⁰ < 12.5 – Base B _T ⁰), it has little affect upon the resultant curve.

Figure 7b shows the percentage of SNIa sample galaxies that have a SB _IR greater than or equal to a given SB_IR. For comparison, we have also shown the curves for the Base sample as well as the Base sample weighted by the SNIa Hubble type distribution. In addition, we have also shown the curve for the modified Base sample that only includes those galaxies that have B_T ^o brighter than 12.5. Figure 7c shows the corresponding plot for SN II sample galaxies.

From figure 7b, we can see that, for SNIa galaxies, the fall off in the number density of data points with increasing SB_IR is close to what you would expect if the SNIa sample galaxies were randomly selected from the Base sample i.e. there appears to be no strong preference for SNIa sample galaxies to either avoid or prefer galaxies with increasing SB_IR .

In contrast, figure 7c shows that the fall off in the number density of data points with increasing SB_IR for SNII galaxies is slower than expected i.e. SNII sample galaxies appear to be preferentially found in galaxies with high SB _IR . This is in agreement with our general impressions of the distribution of the data points in figures 5a and 5b.

Thus, our data shows that there is a population of late type spiral galaxies with high current to integrated star formation rates that are preferentially producing SNII but not SNIa. The preference for SNII, but not SNIa, to appear in galaxies with high current to integrated star formation rates supports our original contention that the enhanced star formation rates in late type spirals are caused by bursts.

Having established this very important result we can now use the SN sample galaxies to study the general properties of the galaxies that are currently experiencing burst activity. . Note: In order to highlight the fact that spirals with high current to integrated star formation rates are undergoing bursts (i.e. they preferentially producing SNII but not SNIa), we have designated all of the galaxies with SB_IR > 35.0 as Bursts and all galaxies with SB_IR < 35.0 as Non-Bursts.