diff --git a/tz.cpp b/tz.cpp
index 3c94ed2..c7b8999 100644
--- a/tz.cpp
+++ b/tz.cpp
@@ -62,6 +62,7 @@
 
 #ifdef _WIN32
 #include <Windows.h>
+#include <io.h>
 #else
 #include <unistd.h>
 #endif
@@ -74,6 +75,15 @@ static const char folder_delimiter =
 '/';
 #endif
 
+static bool file_exists(const std::string& filename)
+{
+#ifdef _WIN32
+    return ::_access(filename.c_str(), 0) == 0;
+#else
+    return ::access(filename.c_str(), F_OK) == 0;
+#endif
+}
+
 #ifdef _WIN32
 // Win32 support requires calling OS functions.
 // This routine maps OS error codes to readable text strngs.
@@ -92,8 +102,8 @@ static std::string get_win32_message(DWORD error_code)
     auto result = FormatMessageA(
         FORMAT_MESSAGE_ALLOCATE_BUFFER | FORMAT_MESSAGE_FROM_SYSTEM
                                        | FORMAT_MESSAGE_IGNORE_INSERTS,
-        NULL, error_code, MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT),
-        reinterpret_cast<char*>(&msg), 0, NULL );
+        nullptr, error_code, MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT),
+        reinterpret_cast<char*>(&msg), 0, nullptr );
     std::unique_ptr<char[], free_message> message_buffer(msg);
     if (result == 0) // If there is no error message, still give the code.
     {
@@ -159,7 +169,7 @@ namespace // Put types in an anonymous name space.
     private:
         // Note there is no value documented to be an invalid handle value.
         // Not NULL nor INVALID_HANDLE_VALUE. We must rely on is_open.
-        HKEY m_key = NULL;
+        HKEY m_key = nullptr;
         bool m_is_open = false;
     public:
         HKEY handle()
@@ -187,7 +197,7 @@ namespace // Put types in an anonymous name space.
                 if (result == ERROR_SUCCESS)
                 {
                     m_is_open = false;
-                    m_key = NULL;
+                    m_key = nullptr;
                 }
                 return result;
             }
@@ -245,13 +255,8 @@ static inline size_t countof(T(&arr)[N])
 
 // This function returns an exhaustive list of time zone information
 // from the Windows registry.
-// The routine tries to to obtain as much information as possible despite errors.
-// If there is an error with any key, it is silently ignored to move on to the next.
-// We don't have a logger to log such errors and it might be disruptive to log anyway.
-// We don't want the whole database of information disrupted just because
-// one record in it can't be read.
-// The expectation is that the errors will eventually manifest to the
-// caller as a missing time zone which they will need to investigate.
+// The routine tries to load as many time zone entries as possible despite errors.
+// We don't want to fail to load the whole database just because one record can't be read.
 
 static void get_windows_timezone_info(std::vector<timezone_info>& tz_list)
 {
@@ -1858,6 +1863,15 @@ init_tzdb()
     std::string line;
     bool continue_zone = false;
     TZ_DB db;
+
+    if (!file_exists(install))
+    {
+        std::string msg = "Timezone database not found at \"";
+        msg += install;
+        msg += "\"";
+        throw std::runtime_error(msg);
+    }
+
     for (const auto& filename : files)
     {
         std::ifstream infile(path + filename);
diff --git a/tz.h b/tz.h
index a1e2bd4..96e3a29 100644
--- a/tz.h
+++ b/tz.h
@@ -33,18 +33,15 @@
 The notion of "current timezone" is something the operating system is expected
 to "just know". How it knows this is system specific. It's often a value
 set by the user at OS intallation time and recorded by the OS somewhere.
-On Linux and Mac systems. the current timezone name is obtained by looking at
+On Linux and Mac systems the current timezone name is obtained by looking at
 the name or contents of a particular file on disk.
 On Windows the current timzeone name comes from the registry.
-But however the name is obtained there is no guarante
-that the "native" current timezone name obtained in this way
+In either method, there is no guarantee that the "native" current timezone name obtained
 will match any of the "Standard" names in this library's "database".
 On Linux, the names usually do seem to match so mapping functions to map from
-native to "Standard" are typically NOT required.
-On Windows, the names are never Standard mapping is always required.
-One should not equate the mapping process with Windows.
-Windows is just currently the only client of them.
-Technically any OS may required mapping.
+native to "Standard" are typically not required.
+On Windows, the names are never "Standard" so mapping is always required.
+Technically any OS may use the mapping process but currently only Windows does use it.
 */
 
 #ifdef _WIN32