diff --git a/README.md b/README.md
index bb7ef04..e2a061f 100644
--- a/README.md
+++ b/README.md
@@ -9,7 +9,7 @@ To connect to the hash circuit, see `spec/hash-table.md`.
 
 The circuit code can be implied with field which have satisified `Hashable` trait and currently only `poseidon-circuit::Bn256Fr` (the alias of `halo2_proofs::halo2curves::bn256::Fr`) has satisified this trait.
 
-The circuit type under `hash::HashCircuit` prove poseidon hash progress base on permutation with 3 fields and a 2 fields rate. You also need to set a fixed step size for proving message hashing with variable length. A message has to be complied with an initial capacity size and for each sponge step the capacity would be substracted by the fixed step size. In the final step the capacity has to be equal or less than the fixed step.
+The circuit type under `hash::HashCircuit` prove poseidon hash progress base on permutation with 3 fields and a 2 fields rate. You also need to set a fixed step size for proving message hashing with variable length. A message has to be complied with an initial capacity size and for each sponge step the capacity would be subtracted by the fixed step size. In the final step the capacity has to be equal or less than the fixed step.
 
 For example, when we hashing a message with 19 fields:
 
diff --git a/spec/hash-table.md b/spec/hash-table.md
index cc962e6..2c48b04 100644
--- a/spec/hash-table.md
+++ b/spec/hash-table.md
@@ -51,7 +51,7 @@ Specifically, `STEP = 32`.
 
 ### Legacy scheme
 
-The initial capcity before `scroll-dev-0215` is not multiplied with the domain mark (`2^64`) so they would get different hash results in *var-len* mode. If we wish
+The initial capacity before `scroll-dev-0215` is not multiplied with the domain mark (`2^64`) so they would get different hash results in *var-len* mode. If we wish
 to apply the code work under the legacy schema, use `legacy` feature
 
 ### Head flag