diff --git a/src/ascii_char.rs b/src/ascii_char.rs
index 39f2ceb..d2803f8 100644
--- a/src/ascii_char.rs
+++ b/src/ascii_char.rs
@@ -468,15 +468,15 @@ impl AsciiChar {
     ///
     /// # Panics
     ///
-    /// Radixes greater than 36 are not supported and will result in a panic.
+    /// Panics if given a radix smaller than 2 or larger than 36.
     #[must_use]
     pub fn is_digit(self, radix: u32) -> bool {
         match (self as u8, radix) {
-            (b'0'..=b'9', 0..=36) => u32::from(self as u8 - b'0') < radix,
+            (b'0'..=b'9', 2..=36) => u32::from(self as u8 - b'0') < radix,
             (b'a'..=b'z', 11..=36) => u32::from(self as u8 - b'a') < radix - 10,
             (b'A'..=b'Z', 11..=36) => u32::from(self as u8 - b'A') < radix - 10,
-            (_, 0..=36) => false,
-            (_, _) => panic!("radixes greater than 36 are not supported"),
+            (_, 2..=36) => false,
+            (_, _) => panic!("radix must be in the range 2 to 36 inclusive"),
         }
     }
 
@@ -704,7 +704,7 @@ impl AsciiChar {
             //         the unprintable characters from 0x0 to 0x1f, ordered correctly.
             //         As `b` is guaranteed to be within 0x0 to 0x1f, the conversion represents a
             //         valid character.
-            b @ 0x0..=0x1f => unsafe { mem::transmute('␀' as u32 + b as u32) },
+            b @ 0x0..=0x1f => unsafe { char::from_u32_unchecked('␀' as u32 + b as u32) },
 
             // 0x7f (delete) has it's own character at codepoint 0x2420, not 0x247f, so it is special
             // cased to return it's character
@@ -794,8 +794,8 @@ macro_rules! impl_into_partial_eq_ord {
     ($wider:ty, $to_wider:expr) => {
         impl From<AsciiChar> for $wider {
             #[inline]
-            fn from(a: AsciiChar) -> $wider {
-                $to_wider(a)
+            fn from(chr: AsciiChar) -> $wider {
+                $to_wider(chr)
             }
         }
         impl PartialEq<$wider> for AsciiChar {
@@ -1050,15 +1050,22 @@ mod tests {
 
     #[test]
     fn is_digit_strange_radixes() {
-        assert_eq!(AsciiChar::_0.is_digit(0), '0'.is_digit(0));
-        assert_eq!(AsciiChar::_0.is_digit(1), '0'.is_digit(1));
+        assert_eq!(AsciiChar::_0.is_digit(2), '0'.is_digit(2));
+        assert_eq!(AsciiChar::_1.is_digit(2), '1'.is_digit(2));
+        assert_eq!(AsciiChar::_2.is_digit(2), '2'.is_digit(2));
         assert_eq!(AsciiChar::_5.is_digit(5), '5'.is_digit(5));
         assert_eq!(AsciiChar::z.is_digit(35), 'z'.is_digit(35));
     }
 
     #[test]
     #[should_panic]
-    fn is_digit_bad_radix() {
+    fn is_digit_bad_radix_small() {
+        let _ = AsciiChar::_0.is_digit(1);
+    }
+
+    #[test]
+    #[should_panic]
+    fn is_digit_bad_radix_big() {
         let _ = AsciiChar::_7.is_digit(37);
     }
 
diff --git a/src/ascii_str.rs b/src/ascii_str.rs
index bfdf54e..3568e85 100644
--- a/src/ascii_str.rs
+++ b/src/ascii_str.rs
@@ -226,7 +226,7 @@ impl AsciiStr {
     /// Returns an iterator over the characters of the `AsciiStr`.
     #[inline]
     #[must_use]
-    pub fn chars(&self) -> Chars {
+    pub fn chars(&self) -> Chars<'_> {
         Chars(self.slice.iter())
     }
 
@@ -234,7 +234,7 @@ impl AsciiStr {
     /// value of each `AsciiChar`.
     #[inline]
     #[must_use]
-    pub fn chars_mut(&mut self) -> CharsMut {
+    pub fn chars_mut(&mut self) -> CharsMut<'_> {
         CharsMut(self.slice.iter_mut())
     }