Mean performances in M2 generation
 
Seedling height registered the highest value in the treatments having a higher concentration (0.20% EMS) in both the varieties (11.6 cm in 
Kashi Nandini and 11.0 cm in 
Kashi Uday), which were above both controls (11.4 cm and 10.9 cm in 
Kashi Nandini and 
Kashi Uday, respectively and was followed by the treatment of 0.15% (9.0 cm in 
Kashi Nandini and 9.2 cm in 
Kashi Uday). In comparison, the lowest value was at 0.10% EMS (8.8 cm in 
Kashi Nandini and 8.9 cm in 
Kashi Uday). These results were in tune with the earlier findings of 
Srivastava et al., (2008) and 
Govardhan and Lal (2013). Days to 50% flowering and maturity in M2 generation showed a delay in all the mutagen treated populations of both varieties (Table 1). The maximum delay in days to 50% flowering and maturity was at 0.20% EMS in both the varieties and the minimum was in controls. The decreasing trend for these two characters with higher concentrations of EMS also agreed with previous findings of 
Ahire (2008), 
Singh et al., (1999) and 
Waghmare and Mehra (2000) also reported similar results. Plant height at maturity showed the highest value at a concentration of 0.20% EMS in both the varieties (60.50 cm in 
Kashi Nandini and 60.1 cm in 
Kashi Uday). The lowest plant height was at 0.10 and 0.15% EMS in both varieties (Table 2), suggesting an increase in plant height at higher concentrations, while a decrease at lower concentrations of EMS in M
2 generation, as also reported by 
Wani and Anis (2001) and 
S.M. Sangle (2020) in Pigeon pea.
 
 
       
The fertile branches were the highest for 0.2% EMS in both varieties, higher than the controls (3.6 in 
Kashi Nandini and 3.7 in 
Kashi Uday, Table 2 and3), followed by 0.15% EMS (3.5 and 3.6 in 
Kashi Nandini and 
Kashi Uday, respectively). The number of fertile branches/plant was the lowest for 0.10% EMS. The results were similar to 
Amjad et al., (1993) in pea variety Green Feast and 
Sadashiv and Kondiram (2012) in horse gram and 
Wani and Anis (2001). The pods/plant and seeds/pod at 0.20% EMS in both varieties significantly differ from the controls in having higher values. The treatments with 0.10 and 0.15% EMS showed low values for these characters. These results were in agreement with 
Savant (2008) in sesame, 
Khan and Praveen (2004) and 
Waghmare and Mehra (2000). Seed yield/plant and 100-seed weights had the highest values at 0.20% EMS in both varieties, followed by 0.10 and 0.15% EMS. These findings were similar to the earlier reports of 
Wani and Anis (2001), 
Pawar et al., (1979), Singh (1973), 
Biradar (2004) and 
Dahiya (1977), 
Patil and Wakode (2011) and 
Wani (2011).
 
 
Induced genetic variability in M2 generation
 
For the improvement of any character, breeders have to select superior individuals based on phenotypic expression.  Phenotype based selection may sometimes be misleading because character development is the result of heritable and non-heritable components. The coefficient of variation is the best tool to measure the relative magnitude of variation among the characters and predicts the variability in the population under study, necessitating partitioning the overall variability into its heritable and non-heritable components. The variability parameters GCV, PCV, heritability and genetic advance for yield and its components are essential to know the clear picture of variability created for different characters. 
       
The estimates of genotypic coefficient of variation (GCV) and phenotypic coefficient of variation (PCV), heritability in a broad sense (h
2b
s) and expected genetic advance (GA) as per cent of the mean for all the characters revealed that the phenotypic coefficient of variation (PCV) was higher than genotypic coefficient of variation (GCV) for all the characters (Fig 1 and 2). Genotypic coefficient of variation (GCV) provides a means to study the genetic variability generated in quantitative characters 
(Johnson et al., 1955). It indirectly measures the environmental influence on the inheritance pattern of yield attributes. Heritability is a good index of the transmission of characters from parents to their offspring. Genetic advance gives the extent of genetic gain. The genetic advance shows the extent of genetic gain that could be expected through selections in character to be improved (
Burton, 1952; 
Johnson et al., 1955).
       
 
 
A comparison of GCV estimates in 
Kashi Nandini and 
Kashi Uday revealed the existence of the highest genetic variation for seedling height, fertile branches/plant and pods/plant at 0.20 and 0.15% EMS (Table 2 and 3), suggesting the scope of selection for these traits for isolating desirable genotypes in garden pea, as also reported by 
Khan and Wani (2006), 
Kumar and Dubey (2001) and 
Kumar (2009) in pea. The results closely agreed with the findings of 
Waghmare and Mehra (2000), 
Muduli and Misra (2008), 
Mensah and Obadoni (2007), 
Khan and Wani (2006), 
Kumar and Dubey (2001) and 
Kumar (2009) in pea. Similarly, the estimates of PCV were also high for seedling height, fertile branches/plant in 
Kashi Nandini, while in 
Kashi Uday; PCV was high for fertile branches/plant, pods/plant at 0.10 and 0.15% EMS (Table 2 and 3). These findings indicated the role of both the genotype and environment for the expression of these characters. These results aligned with the findings of 
Meshram et al., (2014), Sarma et al., (2022), Arul balachandran and Mullainathan (2009). The phenotypic variation was low for the remaining characters (
Wani and Anis, 2014; 
Wani 2011).
       
The highest heritability estimates in the broad sense (h
2b
s) were for seeds/pod and seedling height in 
Kashi Nandini at 0.10% EMS. In 
Kashi Uday, the heritability was high for plant height at maturity at 0.10% EMS and seeds/pod at 0.20% EMS. Moderate estimates recorded for the seedling height in 
Kashi Nandini, seeds/ pod, pods/plant at 0.10 and 0.15% EMS (Fig 1 and 2) indicated that these characters showed a more significant environmental influence and, thus, a more precise evaluation would be necessary for the improvement of these characters. These results had the support of 
Kumar (2009). In the present experiment, heritability in a broad sense was high in treated populations than in control in M
2 generation, which aligned with 
Wani  (2011).
       
As per cent mean, the expected genetic advance is a valuable criterion in a plant breeding programme. Heritability alone is not ideal for predicting the genetic progress that would result from selecting the best individuals from a population. When estimated together, genetic advance and heritability are more helpful in predicting the selection response as high heritability in concurrence with high genetic advance reflects the preponderance of additive gene action 
(Johnson et al., 1955).  As per cent of the mean, the genetic advance was the highest for seedling height at 0.10 and 0.15% EMS in 
Kashi Nandini. The estimates were moderate for seeds/pod with 0.20% EMS, followed by seeds/pod at 0.10% EMS and seedling height at 0.20% EMS in 
Kashi Nandini (Table 2). In 
Kashi Uday, the highest genetic advance was for fertile branches at 0.20% EMS, pods/plant at 0.20% EMS and plant height at 0.10% EMS (Table 3). Moderate estimates were recorded for the characters’ plant height at maturity, followed by pods/plant and fertile branches at 0.15% EMS. The estimates were low for the remaining characters (Fig 1 and 2). The results conformed to 
Khan and Praveen (2004) in green gram, 
Muduli and Misra (2008), 
Mensah and Obadoni (2007) and 
Kumar and Dubey (2001) and 
Wani. A.A. (2011).
 
Desirable mutants observed in M2 generation
 
The present study identified different types of desirable mutants for various quantitative traits in the M
2 generation (Table 4). These mutants showed significant differences over the controls of both the varieties (
Kashi Nandini and 
Kashi Uday, Fig 3 and 4). Desirable mutant plants showed superiority over the controls for fertile branches, pods/plant, seeds/pod and seed yield/plant in both the varieties, as also reported by Tabti 
et  al., (2018), 
Khan et al., (2018), Dhilan et al., (2018), Sharma et al., (2010).