Please analyze the running time of the following variants of merge sort.

$($a$)$Divide the original input array constant times first $($e$.$g$)$ we divide the array $22$ times$)$; then sort the subarrays by using insertion sort. Finally, merge all sorted subarrays recursively to get the sorted array. Dividing the input array a constant times means it stops dividing somewhere even it is not the base case.

$($b$)$ Divide the input array unitl each subarray has constant c $>2$ numbers; sort those subarrays with inserstion sort; merge them recursively.

$($c$)$ Divide the array as normal until each subarray has only one number; mege these subarrays one by one to obtain the whole sorted array.